Upload folder using huggingface_hub
Browse files- .gitattributes +2 -0
- LICENSE +151 -0
- NOTICE +3 -0
- config.json +199 -0
- configuration_intern_vit.py +120 -0
- configuration_qianfanvl_chat.py +84 -0
- conversation.py +358 -0
- example/scene_ocr.png +3 -0
- generation_config.json +6 -0
- model-00001-of-00004.safetensors +3 -0
- model-00002-of-00004.safetensors +3 -0
- model-00003-of-00004.safetensors +3 -0
- model-00004-of-00004.safetensors +3 -0
- model.safetensors.index.json +676 -0
- modeling_intern_vit.py +428 -0
- modeling_qianfanvl_chat.py +394 -0
- special_tokens_map.json +17 -0
- tokenizer.json +3 -0
- tokenizer_config.json +1120 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
example/scene_ocr.png filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
LICENSE
ADDED
|
@@ -0,0 +1,151 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
SPDX-License-Identifier: MIT AND LicenseRef-Meta-Llama-3.1-Community
|
| 2 |
+
|
| 3 |
+
Composite License: MIT (for Original Contributions) + Llama 3.1 Community License (for Llama Materials)
|
| 4 |
+
|
| 5 |
+
=== Section A — MIT License (applies to Original Contributions by the Project) ===
|
| 6 |
+
|
| 7 |
+
MIT License
|
| 8 |
+
|
| 9 |
+
Copyright (c) 2025 Qianfan
|
| 10 |
+
|
| 11 |
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
| 12 |
+
of this software and associated documentation files (the "Software"), to deal
|
| 13 |
+
in the Software without restriction, including without limitation the rights
|
| 14 |
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
| 15 |
+
copies of the Software, and to permit persons to whom the Software is
|
| 16 |
+
furnished to do so, subject to the following conditions:
|
| 17 |
+
|
| 18 |
+
The above copyright notice and this permission notice shall be included in all
|
| 19 |
+
copies or substantial portions of the Software.
|
| 20 |
+
|
| 21 |
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
| 22 |
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
| 23 |
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
| 24 |
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
| 25 |
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
| 26 |
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
| 27 |
+
SOFTWARE.
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
=== Section B — Llama 3.1 Community License (applies to any Meta Llama 3.1 materials included here or derivatives thereof) ===
|
| 31 |
+
|
| 32 |
+
LLAMA 3.1 COMMUNITY LICENSE AGREEMENT
|
| 33 |
+
Llama 3.1 Version Release Date: July 23, 2024
|
| 34 |
+
|
| 35 |
+
“Agreement” means the terms and conditions for use, reproduction, distribution and modification of the
|
| 36 |
+
Llama Materials set forth herein.
|
| 37 |
+
|
| 38 |
+
“Documentation” means the specifications, manuals and documentation accompanying Llama 3.1
|
| 39 |
+
distributed by Meta at https://llama.meta.com/doc/overview.
|
| 40 |
+
|
| 41 |
+
“Licensee” or “you” means you, or your employer or any other person or entity (if you are entering into
|
| 42 |
+
this Agreement on such person or entity’s behalf), of the age required under applicable laws, rules or
|
| 43 |
+
regulations to provide legal consent and that has legal authority to bind your employer or such other
|
| 44 |
+
person or entity if you are entering in this Agreement on their behalf.
|
| 45 |
+
|
| 46 |
+
“Llama 3.1” means the foundational large language models and software and algorithms, including
|
| 47 |
+
machine-learning model code, trained model weights, inference-enabling code, training-enabling code,
|
| 48 |
+
fine-tuning enabling code and other elements of the foregoing distributed by Meta at
|
| 49 |
+
https://llama.meta.com/llama-downloads.
|
| 50 |
+
|
| 51 |
+
“Llama Materials” means, collectively, Meta’s proprietary Llama 3.1 and Documentation (and any
|
| 52 |
+
portion thereof) made available under this Agreement.
|
| 53 |
+
|
| 54 |
+
“Meta” or “we” means Meta Platforms Ireland Limited (if you are located in or, if you are an entity, your
|
| 55 |
+
principal place of business is in the EEA or Switzerland) and Meta Platforms, Inc. (if you are located
|
| 56 |
+
outside of the EEA or Switzerland).
|
| 57 |
+
|
| 58 |
+
By clicking “I Accept” below or by using or distributing any portion or element of the Llama Materials,
|
| 59 |
+
you agree to be bound by this Agreement.
|
| 60 |
+
|
| 61 |
+
1. License Rights and Redistribution.
|
| 62 |
+
|
| 63 |
+
a. Grant of Rights. You are granted a non-exclusive, worldwide, non-transferable and royalty-free
|
| 64 |
+
limited license under Meta’s intellectual property or other rights owned by Meta embodied in the Llama
|
| 65 |
+
Materials to use, reproduce, distribute, copy, create derivative works of, and make modifications to the
|
| 66 |
+
Llama Materials.
|
| 67 |
+
|
| 68 |
+
b. Redistribution and Use.
|
| 69 |
+
|
| 70 |
+
i. If you distribute or make available the Llama Materials (or any derivative works
|
| 71 |
+
thereof), or a product or service (including another AI model) that contains any of them, you shall (A)
|
| 72 |
+
provide a copy of this Agreement with any such Llama Materials; and (B) prominently display “Built with
|
| 73 |
+
Llama” on a related website, user interface, blogpost, about page, or product documentation. If you use
|
| 74 |
+
the Llama Materials or any outputs or results of the Llama Materials to create, train, fine tune, or
|
| 75 |
+
otherwise improve an AI model, which is distributed or made available, you shall also include “Llama” at
|
| 76 |
+
the beginning of any such AI model name.
|
| 77 |
+
|
| 78 |
+
ii. If you receive Llama Materials, or any derivative works thereof, from a Licensee as part
|
| 79 |
+
of an integrated end user product, then Section 2 of this Agreement will not apply to you.
|
| 80 |
+
|
| 81 |
+
iii. You must retain in all copies of the Llama Materials that you distribute the following
|
| 82 |
+
attribution notice within a “Notice” text file distributed as a part of such copies: “Llama 3.1 is
|
| 83 |
+
licensed under the Llama 3.1 Community License, Copyright © Meta Platforms, Inc. All Rights
|
| 84 |
+
Reserved.”
|
| 85 |
+
|
| 86 |
+
iv. Your use of the Llama Materials must comply with applicable laws and regulations
|
| 87 |
+
(including trade compliance laws and regulations) and adhere to the Acceptable Use Policy for the Llama
|
| 88 |
+
Materials (available at https://llama.meta.com/llama3_1/use-policy), which is hereby incorporated by
|
| 89 |
+
reference into this Agreement.
|
| 90 |
+
|
| 91 |
+
2. Additional Commercial Terms. If, on the Llama 3.1 version release date, the monthly active users
|
| 92 |
+
of the products or services made available by or for Licensee, or Licensee’s affiliates, is greater than 700
|
| 93 |
+
million monthly active users in the preceding calendar month, you must request a license from Meta,
|
| 94 |
+
which Meta may grant to you in its sole discretion, and you are not authorized to exercise any of the
|
| 95 |
+
rights under this Agreement unless or until Meta otherwise expressly grants you such rights.
|
| 96 |
+
|
| 97 |
+
3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA MATERIALS AND ANY
|
| 98 |
+
OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF
|
| 99 |
+
ANY KIND, AND META DISCLAIMS ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED,
|
| 100 |
+
INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT,
|
| 101 |
+
MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR
|
| 102 |
+
DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE LLAMA MATERIALS AND
|
| 103 |
+
ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND
|
| 104 |
+
RESULTS.
|
| 105 |
+
|
| 106 |
+
4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF
|
| 107 |
+
LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING
|
| 108 |
+
OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL,
|
| 109 |
+
INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF META OR ITS AFFILIATES HAVE BEEN ADVISED
|
| 110 |
+
OF THE POSSIBILITY OF ANY OF THE FOREGOING.
|
| 111 |
+
|
| 112 |
+
5. Intellectual Property.
|
| 113 |
+
|
| 114 |
+
a. No trademark licenses are granted under this Agreement, and in connection with the Llama
|
| 115 |
+
Materials, neither Meta nor Licensee may use any name or mark owned by or associated with the other
|
| 116 |
+
or any of its affiliates, except as required for reasonable and customary use in describing and
|
| 117 |
+
redistributing the Llama Materials or as set forth in this Section 5(a). Meta hereby grants you a license to
|
| 118 |
+
use “Llama” (the “Mark”) solely as required to comply with the last sentence of Section 1.b.i. You will
|
| 119 |
+
comply with Meta’s brand guidelines (currently accessible at
|
| 120 |
+
https://about.meta.com/brand/resources/meta/company-brand/ ). All goodwill arising out of your use
|
| 121 |
+
of the Mark will inure to the benefit of Meta.
|
| 122 |
+
|
| 123 |
+
b. Subject to Meta’s ownership of Llama Materials and derivatives made by or for Meta, with
|
| 124 |
+
respect to any derivative works and modifications of the Llama Materials that are made by you, as
|
| 125 |
+
between you and Meta, you are and will be the owner of such derivative works and modifications.
|
| 126 |
+
|
| 127 |
+
c. If you institute litigation or other proceedings against Meta or any entity (including a
|
| 128 |
+
cross-claim or counterclaim in a lawsuit) alleging that the Llama Materials or Llama 3.1 outputs or
|
| 129 |
+
results, or any portion of any of the foregoing, constitutes infringement of intellectual property or other
|
| 130 |
+
rights owned or licensable by you, then any licenses granted to you under this Agreement shall
|
| 131 |
+
terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold
|
| 132 |
+
harmless Meta from and against any claim by any third party arising out of or related to your use or
|
| 133 |
+
distribution of the Llama Materials.
|
| 134 |
+
|
| 135 |
+
6. Term and Termination. The term of this Agreement will commence upon your acceptance of this
|
| 136 |
+
Agreement or access to the Llama Materials and will continue in full force and effect until terminated in
|
| 137 |
+
accordance with the terms and conditions herein. Meta may terminate this Agreement if you are in
|
| 138 |
+
breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete
|
| 139 |
+
and cease use of the Llama Materials. Sections 3, 4 and 7 shall survive the termination of this
|
| 140 |
+
Agreement.
|
| 141 |
+
|
| 142 |
+
7. Governing Law and Jurisdiction. This Agreement will be governed and construed under the laws of
|
| 143 |
+
the State of California without regard to choice of law principles, and the UN Convention on Contracts
|
| 144 |
+
for the International Sale of Goods does not apply to this Agreement. The courts of California shall have
|
| 145 |
+
exclusive jurisdiction of any dispute arising out of this Agreement.
|
| 146 |
+
|
| 147 |
+
|
| 148 |
+
=== Scope Clarification (Non‑operative summary) ===
|
| 149 |
+
- Section A (MIT) covers only the Project’s original contributions authored by Qianfan.
|
| 150 |
+
- Section B (Llama 3.1 Community License) governs any included Llama Materials and any derivatives thereof (e.g., fine‑tuned weights).
|
| 151 |
+
- In the event of any conflict, the applicable license for the relevant component controls (MIT for original contributions; Llama 3.1 for Llama Materials).
|
NOTICE
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Llama 3.1 is licensed under the Llama 3.1 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved.
|
| 2 |
+
|
| 3 |
+
Built with Llama
|
config.json
ADDED
|
@@ -0,0 +1,199 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_commit_hash": null,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"QianfanVLChatModel"
|
| 5 |
+
],
|
| 6 |
+
"auto_map": {
|
| 7 |
+
"AutoConfig": "configuration_qianfanvl_chat.QianfanVLChatConfig",
|
| 8 |
+
"AutoModel": "modeling_qianfanvl_chat.QianfanVLChatModel",
|
| 9 |
+
"AutoModelForCausalLM": "modeling_qianfanvl_chat.QianfanVLChatModel"
|
| 10 |
+
},
|
| 11 |
+
"downsample_ratio": 0.5,
|
| 12 |
+
"dynamic_image_size": true,
|
| 13 |
+
"force_image_size": 448,
|
| 14 |
+
"hidden_size": 4096,
|
| 15 |
+
"llm_config": {
|
| 16 |
+
"_name_or_path": "",
|
| 17 |
+
"add_cross_attention": false,
|
| 18 |
+
"architectures": [
|
| 19 |
+
"LlamaForCausalLM"
|
| 20 |
+
],
|
| 21 |
+
"attention_bias": false,
|
| 22 |
+
"attention_dropout": 0.0,
|
| 23 |
+
"attn_implementation": "eager",
|
| 24 |
+
"bad_words_ids": null,
|
| 25 |
+
"begin_suppress_tokens": null,
|
| 26 |
+
"bos_token_id": 181887,
|
| 27 |
+
"chunk_size_feed_forward": 0,
|
| 28 |
+
"cross_attention_hidden_size": null,
|
| 29 |
+
"decoder_start_token_id": null,
|
| 30 |
+
"diversity_penalty": 0.0,
|
| 31 |
+
"do_sample": false,
|
| 32 |
+
"early_stopping": false,
|
| 33 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 34 |
+
"eos_token_id": 181888,
|
| 35 |
+
"exponential_decay_length_penalty": null,
|
| 36 |
+
"finetuning_task": null,
|
| 37 |
+
"forced_bos_token_id": null,
|
| 38 |
+
"forced_eos_token_id": null,
|
| 39 |
+
"hidden_act": "silu",
|
| 40 |
+
"hidden_size": 4096,
|
| 41 |
+
"id2label": {
|
| 42 |
+
"0": "LABEL_0",
|
| 43 |
+
"1": "LABEL_1"
|
| 44 |
+
},
|
| 45 |
+
"initializer_range": 0.02,
|
| 46 |
+
"intermediate_size": 14336,
|
| 47 |
+
"is_decoder": false,
|
| 48 |
+
"is_encoder_decoder": false,
|
| 49 |
+
"label2id": {
|
| 50 |
+
"LABEL_0": 0,
|
| 51 |
+
"LABEL_1": 1
|
| 52 |
+
},
|
| 53 |
+
"length_penalty": 1.0,
|
| 54 |
+
"max_length": 20,
|
| 55 |
+
"max_position_embeddings": 32768,
|
| 56 |
+
"min_length": 0,
|
| 57 |
+
"model_type": "llama",
|
| 58 |
+
"no_repeat_ngram_size": 0,
|
| 59 |
+
"num_attention_heads": 32,
|
| 60 |
+
"num_beam_groups": 1,
|
| 61 |
+
"num_beams": 1,
|
| 62 |
+
"num_hidden_layers": 32,
|
| 63 |
+
"num_key_value_heads": 8,
|
| 64 |
+
"num_return_sequences": 1,
|
| 65 |
+
"output_attentions": false,
|
| 66 |
+
"output_hidden_states": false,
|
| 67 |
+
"output_scores": false,
|
| 68 |
+
"pad_token_id": null,
|
| 69 |
+
"prefix": null,
|
| 70 |
+
"pretraining_tp": 1,
|
| 71 |
+
"problem_type": null,
|
| 72 |
+
"pruned_heads": {},
|
| 73 |
+
"remove_invalid_values": false,
|
| 74 |
+
"repetition_penalty": 1.0,
|
| 75 |
+
"return_dict": true,
|
| 76 |
+
"return_dict_in_generate": false,
|
| 77 |
+
"rms_norm_eps": 1e-05,
|
| 78 |
+
"rope_scaling": null,
|
| 79 |
+
"rope_theta": 5000000.0,
|
| 80 |
+
"sep_token_id": null,
|
| 81 |
+
"suppress_tokens": null,
|
| 82 |
+
"task_specific_params": null,
|
| 83 |
+
"temperature": 1.0,
|
| 84 |
+
"tf_legacy_loss": false,
|
| 85 |
+
"tie_encoder_decoder": false,
|
| 86 |
+
"tie_word_embeddings": false,
|
| 87 |
+
"tokenizer_class": null,
|
| 88 |
+
"top_k": 50,
|
| 89 |
+
"top_p": 1.0,
|
| 90 |
+
"torch_dtype": "bfloat16",
|
| 91 |
+
"torchscript": false,
|
| 92 |
+
"transformers_version": "4.37.2",
|
| 93 |
+
"typical_p": 1.0,
|
| 94 |
+
"use_bfloat16": false,
|
| 95 |
+
"use_cache": true,
|
| 96 |
+
"vocab_size": 182025
|
| 97 |
+
},
|
| 98 |
+
"max_dynamic_patch": 12,
|
| 99 |
+
"min_dynamic_patch": 1,
|
| 100 |
+
"model_type": "qianfanvl_chat",
|
| 101 |
+
"pad2square": false,
|
| 102 |
+
"ps_version": "v2",
|
| 103 |
+
"select_layer": -1,
|
| 104 |
+
"template": "qianfanvl",
|
| 105 |
+
"tie_word_embeddings": false,
|
| 106 |
+
"torch_dtype": "bfloat16",
|
| 107 |
+
"transformers_version": null,
|
| 108 |
+
"use_backbone_lora": 0,
|
| 109 |
+
"use_llm_lora": 0,
|
| 110 |
+
"use_thumbnail": true,
|
| 111 |
+
"vision_config": {
|
| 112 |
+
"_name_or_path": "",
|
| 113 |
+
"add_cross_attention": false,
|
| 114 |
+
"architectures": [
|
| 115 |
+
"InternVisionModel"
|
| 116 |
+
],
|
| 117 |
+
"attention_dropout": 0.0,
|
| 118 |
+
"auto_map": {
|
| 119 |
+
"AutoConfig": "configuration_intern_vit.InternVisionConfig",
|
| 120 |
+
"AutoModel": "modeling_intern_vit.InternVisionModel"
|
| 121 |
+
},
|
| 122 |
+
"bad_words_ids": null,
|
| 123 |
+
"begin_suppress_tokens": null,
|
| 124 |
+
"bos_token_id": null,
|
| 125 |
+
"chunk_size_feed_forward": 0,
|
| 126 |
+
"cross_attention_hidden_size": null,
|
| 127 |
+
"decoder_start_token_id": null,
|
| 128 |
+
"diversity_penalty": 0.0,
|
| 129 |
+
"do_sample": false,
|
| 130 |
+
"drop_path_rate": 0.0,
|
| 131 |
+
"dropout": 0.0,
|
| 132 |
+
"early_stopping": false,
|
| 133 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 134 |
+
"eos_token_id": null,
|
| 135 |
+
"exponential_decay_length_penalty": null,
|
| 136 |
+
"finetuning_task": null,
|
| 137 |
+
"forced_bos_token_id": null,
|
| 138 |
+
"forced_eos_token_id": null,
|
| 139 |
+
"hidden_act": "gelu",
|
| 140 |
+
"hidden_size": 1024,
|
| 141 |
+
"id2label": {
|
| 142 |
+
"0": "LABEL_0",
|
| 143 |
+
"1": "LABEL_1"
|
| 144 |
+
},
|
| 145 |
+
"image_size": 448,
|
| 146 |
+
"initializer_factor": 1.0,
|
| 147 |
+
"initializer_range": 0.02,
|
| 148 |
+
"intermediate_size": 4096,
|
| 149 |
+
"is_decoder": false,
|
| 150 |
+
"is_encoder_decoder": false,
|
| 151 |
+
"label2id": {
|
| 152 |
+
"LABEL_0": 0,
|
| 153 |
+
"LABEL_1": 1
|
| 154 |
+
},
|
| 155 |
+
"layer_norm_eps": 1e-06,
|
| 156 |
+
"length_penalty": 1.0,
|
| 157 |
+
"max_length": 20,
|
| 158 |
+
"min_length": 0,
|
| 159 |
+
"model_type": "intern_vit_6b",
|
| 160 |
+
"no_repeat_ngram_size": 0,
|
| 161 |
+
"norm_type": "layer_norm",
|
| 162 |
+
"num_attention_heads": 16,
|
| 163 |
+
"num_beam_groups": 1,
|
| 164 |
+
"num_beams": 1,
|
| 165 |
+
"num_channels": 3,
|
| 166 |
+
"num_hidden_layers": 24,
|
| 167 |
+
"num_return_sequences": 1,
|
| 168 |
+
"output_attentions": false,
|
| 169 |
+
"output_hidden_states": false,
|
| 170 |
+
"output_scores": false,
|
| 171 |
+
"pad_token_id": null,
|
| 172 |
+
"patch_size": 14,
|
| 173 |
+
"prefix": null,
|
| 174 |
+
"problem_type": null,
|
| 175 |
+
"pruned_heads": {},
|
| 176 |
+
"qk_normalization": false,
|
| 177 |
+
"qkv_bias": true,
|
| 178 |
+
"remove_invalid_values": false,
|
| 179 |
+
"repetition_penalty": 1.0,
|
| 180 |
+
"return_dict": true,
|
| 181 |
+
"return_dict_in_generate": false,
|
| 182 |
+
"sep_token_id": null,
|
| 183 |
+
"suppress_tokens": null,
|
| 184 |
+
"task_specific_params": null,
|
| 185 |
+
"temperature": 1.0,
|
| 186 |
+
"tf_legacy_loss": false,
|
| 187 |
+
"tie_encoder_decoder": false,
|
| 188 |
+
"tie_word_embeddings": true,
|
| 189 |
+
"tokenizer_class": null,
|
| 190 |
+
"top_k": 50,
|
| 191 |
+
"top_p": 1.0,
|
| 192 |
+
"torch_dtype": "bfloat16",
|
| 193 |
+
"torchscript": false,
|
| 194 |
+
"transformers_version": "4.37.2",
|
| 195 |
+
"typical_p": 1.0,
|
| 196 |
+
"use_bfloat16": false,
|
| 197 |
+
"use_flash_attn": false
|
| 198 |
+
}
|
| 199 |
+
}
|
configuration_intern_vit.py
ADDED
|
@@ -0,0 +1,120 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# --------------------------------------------------------
|
| 2 |
+
# InternVL
|
| 3 |
+
# Copyright (c) 2024 OpenGVLab
|
| 4 |
+
# Licensed under The MIT License [see LICENSE for details]
|
| 5 |
+
# --------------------------------------------------------
|
| 6 |
+
|
| 7 |
+
import os
|
| 8 |
+
from typing import Union
|
| 9 |
+
|
| 10 |
+
from transformers.configuration_utils import PretrainedConfig
|
| 11 |
+
from transformers.utils import logging
|
| 12 |
+
|
| 13 |
+
logger = logging.get_logger(__name__)
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
class InternVisionConfig(PretrainedConfig):
|
| 17 |
+
r"""
|
| 18 |
+
This is the configuration class to store the configuration of a [`InternVisionModel`]. It is used to
|
| 19 |
+
instantiate a vision encoder according to the specified arguments, defining the model architecture.
|
| 20 |
+
|
| 21 |
+
Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
|
| 22 |
+
documentation from [`PretrainedConfig`] for more information.
|
| 23 |
+
|
| 24 |
+
Args:
|
| 25 |
+
num_channels (`int`, *optional*, defaults to 3):
|
| 26 |
+
Number of color channels in the input images (e.g., 3 for RGB).
|
| 27 |
+
patch_size (`int`, *optional*, defaults to 14):
|
| 28 |
+
The size (resolution) of each patch.
|
| 29 |
+
image_size (`int`, *optional*, defaults to 224):
|
| 30 |
+
The size (resolution) of each image.
|
| 31 |
+
qkv_bias (`bool`, *optional*, defaults to `False`):
|
| 32 |
+
Whether to add a bias to the queries and values in the self-attention layers.
|
| 33 |
+
hidden_size (`int`, *optional*, defaults to 3200):
|
| 34 |
+
Dimensionality of the encoder layers and the pooler layer.
|
| 35 |
+
num_attention_heads (`int`, *optional*, defaults to 25):
|
| 36 |
+
Number of attention heads for each attention layer in the Transformer encoder.
|
| 37 |
+
intermediate_size (`int`, *optional*, defaults to 12800):
|
| 38 |
+
Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
|
| 39 |
+
qk_normalization (`bool`, *optional*, defaults to `True`):
|
| 40 |
+
Whether to normalize the queries and keys in the self-attention layers.
|
| 41 |
+
num_hidden_layers (`int`, *optional*, defaults to 48):
|
| 42 |
+
Number of hidden layers in the Transformer encoder.
|
| 43 |
+
use_flash_attn (`bool`, *optional*, defaults to `True`):
|
| 44 |
+
Whether to use flash attention mechanism.
|
| 45 |
+
hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`):
|
| 46 |
+
The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
|
| 47 |
+
`"relu"`, `"selu"` and `"gelu_new"` ``"gelu"` are supported.
|
| 48 |
+
layer_norm_eps (`float`, *optional*, defaults to 1e-6):
|
| 49 |
+
The epsilon used by the layer normalization layers.
|
| 50 |
+
dropout (`float`, *optional*, defaults to 0.0):
|
| 51 |
+
The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
|
| 52 |
+
drop_path_rate (`float`, *optional*, defaults to 0.0):
|
| 53 |
+
Dropout rate for stochastic depth.
|
| 54 |
+
attention_dropout (`float`, *optional*, defaults to 0.0):
|
| 55 |
+
The dropout ratio for the attention probabilities.
|
| 56 |
+
initializer_range (`float`, *optional*, defaults to 0.02):
|
| 57 |
+
The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
|
| 58 |
+
initializer_factor (`float`, *optional*, defaults to 0.1):
|
| 59 |
+
A factor for layer scale.
|
| 60 |
+
"""
|
| 61 |
+
|
| 62 |
+
model_type = 'intern_vit_6b'
|
| 63 |
+
|
| 64 |
+
def __init__(
|
| 65 |
+
self,
|
| 66 |
+
num_channels=3,
|
| 67 |
+
patch_size=14,
|
| 68 |
+
image_size=224,
|
| 69 |
+
qkv_bias=False,
|
| 70 |
+
hidden_size=3200,
|
| 71 |
+
num_attention_heads=25,
|
| 72 |
+
intermediate_size=12800,
|
| 73 |
+
qk_normalization=True,
|
| 74 |
+
num_hidden_layers=48,
|
| 75 |
+
use_flash_attn=True,
|
| 76 |
+
hidden_act='gelu',
|
| 77 |
+
norm_type='rms_norm',
|
| 78 |
+
layer_norm_eps=1e-6,
|
| 79 |
+
dropout=0.0,
|
| 80 |
+
drop_path_rate=0.0,
|
| 81 |
+
attention_dropout=0.0,
|
| 82 |
+
initializer_range=0.02,
|
| 83 |
+
initializer_factor=0.1,
|
| 84 |
+
**kwargs,
|
| 85 |
+
):
|
| 86 |
+
super().__init__(**kwargs)
|
| 87 |
+
|
| 88 |
+
self.hidden_size = hidden_size
|
| 89 |
+
self.intermediate_size = intermediate_size
|
| 90 |
+
self.dropout = dropout
|
| 91 |
+
self.drop_path_rate = drop_path_rate
|
| 92 |
+
self.num_hidden_layers = num_hidden_layers
|
| 93 |
+
self.num_attention_heads = num_attention_heads
|
| 94 |
+
self.num_channels = num_channels
|
| 95 |
+
self.patch_size = patch_size
|
| 96 |
+
self.image_size = image_size
|
| 97 |
+
self.initializer_range = initializer_range
|
| 98 |
+
self.initializer_factor = initializer_factor
|
| 99 |
+
self.attention_dropout = attention_dropout
|
| 100 |
+
self.layer_norm_eps = layer_norm_eps
|
| 101 |
+
self.hidden_act = hidden_act
|
| 102 |
+
self.norm_type = norm_type
|
| 103 |
+
self.qkv_bias = qkv_bias
|
| 104 |
+
self.qk_normalization = qk_normalization
|
| 105 |
+
self.use_flash_attn = use_flash_attn
|
| 106 |
+
|
| 107 |
+
@classmethod
|
| 108 |
+
def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> 'PretrainedConfig':
|
| 109 |
+
config_dict, kwargs = cls.get_config_dict(pretrained_model_name_or_path, **kwargs)
|
| 110 |
+
|
| 111 |
+
if 'vision_config' in config_dict:
|
| 112 |
+
config_dict = config_dict['vision_config']
|
| 113 |
+
|
| 114 |
+
if 'model_type' in config_dict and hasattr(cls, 'model_type') and config_dict['model_type'] != cls.model_type:
|
| 115 |
+
logger.warning(
|
| 116 |
+
f"You are using a model of type {config_dict['model_type']} to instantiate a model of type "
|
| 117 |
+
f'{cls.model_type}. This is not supported for all configurations of models and can yield errors.'
|
| 118 |
+
)
|
| 119 |
+
|
| 120 |
+
return cls.from_dict(config_dict, **kwargs)
|
configuration_qianfanvl_chat.py
ADDED
|
@@ -0,0 +1,84 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Copyright (c) 2025 Qianfan
|
| 2 |
+
# Licensed under the MIT License. See LICENSE file in the project root for full license information.
|
| 3 |
+
import copy
|
| 4 |
+
|
| 5 |
+
from transformers import AutoConfig, LlamaConfig
|
| 6 |
+
from transformers.configuration_utils import PretrainedConfig
|
| 7 |
+
from transformers.utils import logging
|
| 8 |
+
|
| 9 |
+
from .configuration_intern_vit import InternVisionConfig
|
| 10 |
+
|
| 11 |
+
logger = logging.get_logger(__name__)
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
class QianfanVLChatConfig(PretrainedConfig):
|
| 15 |
+
model_type = 'qianfanvl_chat'
|
| 16 |
+
is_composition = True
|
| 17 |
+
|
| 18 |
+
def __init__(
|
| 19 |
+
self,
|
| 20 |
+
vision_config=None,
|
| 21 |
+
llm_config=None,
|
| 22 |
+
use_backbone_lora=0,
|
| 23 |
+
use_llm_lora=0,
|
| 24 |
+
pad2square=False,
|
| 25 |
+
select_layer=-1,
|
| 26 |
+
force_image_size=None,
|
| 27 |
+
downsample_ratio=0.5,
|
| 28 |
+
template=None,
|
| 29 |
+
dynamic_image_size=False,
|
| 30 |
+
use_thumbnail=False,
|
| 31 |
+
ps_version='v1',
|
| 32 |
+
min_dynamic_patch=1,
|
| 33 |
+
max_dynamic_patch=6,
|
| 34 |
+
**kwargs):
|
| 35 |
+
super().__init__(**kwargs)
|
| 36 |
+
if vision_config is None:
|
| 37 |
+
vision_config = {'architectures': ['InternVisionModel']}
|
| 38 |
+
if llm_config is None:
|
| 39 |
+
llm_config = {'architectures': ['LlamaForCausalLM']}
|
| 40 |
+
self.vision_config = InternVisionConfig(**vision_config)
|
| 41 |
+
self.llm_config = LlamaConfig(**llm_config)
|
| 42 |
+
self.text_config = self.llm_config.get_text_config()
|
| 43 |
+
self.use_backbone_lora = use_backbone_lora
|
| 44 |
+
self.use_llm_lora = use_llm_lora
|
| 45 |
+
self.pad2square = pad2square
|
| 46 |
+
self.select_layer = select_layer
|
| 47 |
+
self.force_image_size = force_image_size
|
| 48 |
+
self.downsample_ratio = downsample_ratio
|
| 49 |
+
self.template = template
|
| 50 |
+
self.dynamic_image_size = dynamic_image_size
|
| 51 |
+
self.use_thumbnail = use_thumbnail
|
| 52 |
+
self.ps_version = ps_version # pixel shuffle version
|
| 53 |
+
self.min_dynamic_patch = min_dynamic_patch
|
| 54 |
+
self.max_dynamic_patch = max_dynamic_patch
|
| 55 |
+
|
| 56 |
+
self.hidden_size = self.llm_config.hidden_size
|
| 57 |
+
self.tie_word_embeddings = False
|
| 58 |
+
self.llm_config.tie_word_embeddings = self.tie_word_embeddings
|
| 59 |
+
|
| 60 |
+
def to_dict(self):
|
| 61 |
+
"""
|
| 62 |
+
Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
|
| 63 |
+
|
| 64 |
+
Returns:
|
| 65 |
+
`Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
|
| 66 |
+
"""
|
| 67 |
+
output = copy.deepcopy(self.__dict__)
|
| 68 |
+
output['vision_config'] = self.vision_config.to_dict()
|
| 69 |
+
output['llm_config'] = self.llm_config.to_dict()
|
| 70 |
+
output['text_config'] = self.text_config.to_dict()
|
| 71 |
+
output['model_type'] = self.__class__.model_type
|
| 72 |
+
output['use_backbone_lora'] = self.use_backbone_lora
|
| 73 |
+
output['use_llm_lora'] = self.use_llm_lora
|
| 74 |
+
output['select_layer'] = self.select_layer
|
| 75 |
+
output['force_image_size'] = self.force_image_size
|
| 76 |
+
output['downsample_ratio'] = self.downsample_ratio
|
| 77 |
+
output['template'] = self.template
|
| 78 |
+
output['dynamic_image_size'] = self.dynamic_image_size
|
| 79 |
+
output['use_thumbnail'] = self.use_thumbnail
|
| 80 |
+
output['ps_version'] = self.ps_version
|
| 81 |
+
output['min_dynamic_patch'] = self.min_dynamic_patch
|
| 82 |
+
output['max_dynamic_patch'] = self.max_dynamic_patch
|
| 83 |
+
|
| 84 |
+
return output
|
conversation.py
ADDED
|
@@ -0,0 +1,358 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
"""
|
| 2 |
+
Conversation prompt templates.
|
| 3 |
+
|
| 4 |
+
We kindly request that you import fastchat instead of copying this file if you wish to use it.
|
| 5 |
+
If you have changes in mind, please contribute back so the community can benefit collectively and continue to maintain these valuable templates.
|
| 6 |
+
|
| 7 |
+
Modified from https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py
|
| 8 |
+
"""
|
| 9 |
+
|
| 10 |
+
import dataclasses
|
| 11 |
+
from enum import IntEnum, auto
|
| 12 |
+
from typing import Any, Dict, List, Tuple, Union
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
class SeparatorStyle(IntEnum):
|
| 16 |
+
"""Separator styles."""
|
| 17 |
+
|
| 18 |
+
ADD_COLON_SINGLE = auto()
|
| 19 |
+
ADD_COLON_TWO = auto()
|
| 20 |
+
ADD_COLON_SPACE_SINGLE = auto()
|
| 21 |
+
NO_COLON_SINGLE = auto()
|
| 22 |
+
NO_COLON_TWO = auto()
|
| 23 |
+
ADD_NEW_LINE_SINGLE = auto()
|
| 24 |
+
LLAMA2 = auto()
|
| 25 |
+
CHATGLM = auto()
|
| 26 |
+
CHATML = auto()
|
| 27 |
+
CHATINTERN = auto()
|
| 28 |
+
DOLLY = auto()
|
| 29 |
+
RWKV = auto()
|
| 30 |
+
PHOENIX = auto()
|
| 31 |
+
ROBIN = auto()
|
| 32 |
+
FALCON_CHAT = auto()
|
| 33 |
+
CHATGLM3 = auto()
|
| 34 |
+
INTERNVL_ZH = auto()
|
| 35 |
+
MPT = auto()
|
| 36 |
+
|
| 37 |
+
|
| 38 |
+
@dataclasses.dataclass
|
| 39 |
+
class Conversation:
|
| 40 |
+
"""A class that manages prompt templates and keeps all conversation history."""
|
| 41 |
+
|
| 42 |
+
# The name of this template
|
| 43 |
+
name: str
|
| 44 |
+
# The template of the system prompt
|
| 45 |
+
system_template: str = '{system_message}'
|
| 46 |
+
# The system message
|
| 47 |
+
system_message: str = ''
|
| 48 |
+
# The names of two roles
|
| 49 |
+
roles: Tuple[str] = ('USER', 'ASSISTANT')
|
| 50 |
+
# All messages. Each item is (role, message).
|
| 51 |
+
messages: List[List[str]] = ()
|
| 52 |
+
# The number of few shot examples
|
| 53 |
+
offset: int = 0
|
| 54 |
+
# The separator style and configurations
|
| 55 |
+
sep_style: SeparatorStyle = SeparatorStyle.ADD_COLON_SINGLE
|
| 56 |
+
sep: str = '\n'
|
| 57 |
+
sep2: str = None
|
| 58 |
+
# Stop criteria (the default one is EOS token)
|
| 59 |
+
stop_str: Union[str, List[str]] = None
|
| 60 |
+
# Stops generation if meeting any token in this list
|
| 61 |
+
stop_token_ids: List[int] = None
|
| 62 |
+
|
| 63 |
+
def get_prompt(self) -> str:
|
| 64 |
+
"""Get the prompt for generation."""
|
| 65 |
+
system_prompt = self.system_template.format(system_message=self.system_message)
|
| 66 |
+
if self.sep_style == SeparatorStyle.ADD_COLON_SINGLE:
|
| 67 |
+
ret = system_prompt + self.sep
|
| 68 |
+
for role, message in self.messages:
|
| 69 |
+
if message:
|
| 70 |
+
ret += role + ': ' + message + self.sep
|
| 71 |
+
else:
|
| 72 |
+
ret += role + ':'
|
| 73 |
+
return ret
|
| 74 |
+
elif self.sep_style == SeparatorStyle.ADD_COLON_TWO:
|
| 75 |
+
seps = [self.sep, self.sep2]
|
| 76 |
+
ret = system_prompt + seps[0]
|
| 77 |
+
for i, (role, message) in enumerate(self.messages):
|
| 78 |
+
if message:
|
| 79 |
+
ret += role + ': ' + message + seps[i % 2]
|
| 80 |
+
else:
|
| 81 |
+
ret += role + ':'
|
| 82 |
+
return ret
|
| 83 |
+
elif self.sep_style == SeparatorStyle.ADD_COLON_SPACE_SINGLE:
|
| 84 |
+
ret = system_prompt + self.sep
|
| 85 |
+
for role, message in self.messages:
|
| 86 |
+
if message:
|
| 87 |
+
ret += role + ': ' + message + self.sep
|
| 88 |
+
else:
|
| 89 |
+
ret += role + ': ' # must be end with a space
|
| 90 |
+
return ret
|
| 91 |
+
elif self.sep_style == SeparatorStyle.ADD_NEW_LINE_SINGLE:
|
| 92 |
+
ret = '' if system_prompt == '' else system_prompt + self.sep
|
| 93 |
+
for role, message in self.messages:
|
| 94 |
+
if message:
|
| 95 |
+
ret += role + '\n' + message + self.sep
|
| 96 |
+
else:
|
| 97 |
+
ret += role + '\n'
|
| 98 |
+
return ret
|
| 99 |
+
elif self.sep_style == SeparatorStyle.NO_COLON_SINGLE:
|
| 100 |
+
ret = system_prompt
|
| 101 |
+
for role, message in self.messages:
|
| 102 |
+
if message:
|
| 103 |
+
ret += role + message + self.sep
|
| 104 |
+
else:
|
| 105 |
+
ret += role
|
| 106 |
+
return ret
|
| 107 |
+
elif self.sep_style == SeparatorStyle.NO_COLON_TWO:
|
| 108 |
+
seps = [self.sep, self.sep2]
|
| 109 |
+
ret = system_prompt
|
| 110 |
+
for i, (role, message) in enumerate(self.messages):
|
| 111 |
+
if message:
|
| 112 |
+
ret += role + message + seps[i % 2]
|
| 113 |
+
else:
|
| 114 |
+
ret += role
|
| 115 |
+
return ret
|
| 116 |
+
elif self.sep_style == SeparatorStyle.RWKV:
|
| 117 |
+
ret = system_prompt
|
| 118 |
+
for i, (role, message) in enumerate(self.messages):
|
| 119 |
+
if message:
|
| 120 |
+
ret += (
|
| 121 |
+
role
|
| 122 |
+
+ ': '
|
| 123 |
+
+ message.replace('\r\n', '\n').replace('\n\n', '\n')
|
| 124 |
+
)
|
| 125 |
+
ret += '\n\n'
|
| 126 |
+
else:
|
| 127 |
+
ret += role + ':'
|
| 128 |
+
return ret
|
| 129 |
+
elif self.sep_style == SeparatorStyle.LLAMA2:
|
| 130 |
+
seps = [self.sep, self.sep2]
|
| 131 |
+
if self.system_message:
|
| 132 |
+
ret = system_prompt
|
| 133 |
+
else:
|
| 134 |
+
ret = '[INST] '
|
| 135 |
+
for i, (role, message) in enumerate(self.messages):
|
| 136 |
+
tag = self.roles[i % 2]
|
| 137 |
+
if message:
|
| 138 |
+
if i == 0:
|
| 139 |
+
ret += message + ' '
|
| 140 |
+
else:
|
| 141 |
+
ret += tag + ' ' + message + seps[i % 2]
|
| 142 |
+
else:
|
| 143 |
+
ret += tag
|
| 144 |
+
return ret
|
| 145 |
+
elif self.sep_style == SeparatorStyle.CHATGLM:
|
| 146 |
+
# source: https://huggingface.co/THUDM/chatglm-6b/blob/1d240ba371910e9282298d4592532d7f0f3e9f3e/modeling_chatglm.py#L1302-L1308
|
| 147 |
+
# source2: https://huggingface.co/THUDM/chatglm2-6b/blob/e186c891cf64310ac66ef10a87e6635fa6c2a579/modeling_chatglm.py#L926
|
| 148 |
+
round_add_n = 1 if self.name == 'chatglm2' else 0
|
| 149 |
+
if system_prompt:
|
| 150 |
+
ret = system_prompt + self.sep
|
| 151 |
+
else:
|
| 152 |
+
ret = ''
|
| 153 |
+
|
| 154 |
+
for i, (role, message) in enumerate(self.messages):
|
| 155 |
+
if i % 2 == 0:
|
| 156 |
+
ret += f'[Round {i//2 + round_add_n}]{self.sep}'
|
| 157 |
+
|
| 158 |
+
if message:
|
| 159 |
+
ret += f'{role}：{message}{self.sep}'
|
| 160 |
+
else:
|
| 161 |
+
ret += f'{role}：'
|
| 162 |
+
return ret
|
| 163 |
+
elif self.sep_style == SeparatorStyle.CHATML:
|
| 164 |
+
ret = '' if system_prompt == '' else system_prompt + self.sep + '\n'
|
| 165 |
+
for role, message in self.messages:
|
| 166 |
+
if message:
|
| 167 |
+
ret += role + '\n' + message + self.sep + '\n'
|
| 168 |
+
else:
|
| 169 |
+
ret += role + '\n'
|
| 170 |
+
return ret
|
| 171 |
+
elif self.sep_style == SeparatorStyle.CHATGLM3:
|
| 172 |
+
ret = ''
|
| 173 |
+
if self.system_message:
|
| 174 |
+
ret += system_prompt
|
| 175 |
+
for role, message in self.messages:
|
| 176 |
+
if message:
|
| 177 |
+
ret += role + '\n' + ' ' + message
|
| 178 |
+
else:
|
| 179 |
+
ret += role
|
| 180 |
+
return ret
|
| 181 |
+
elif self.sep_style == SeparatorStyle.CHATINTERN:
|
| 182 |
+
# source: https://huggingface.co/internlm/internlm-chat-7b-8k/blob/bd546fa984b4b0b86958f56bf37f94aa75ab8831/modeling_internlm.py#L771
|
| 183 |
+
seps = [self.sep, self.sep2]
|
| 184 |
+
ret = system_prompt
|
| 185 |
+
for i, (role, message) in enumerate(self.messages):
|
| 186 |
+
# if i % 2 == 0:
|
| 187 |
+
# ret += "<s>"
|
| 188 |
+
if message:
|
| 189 |
+
ret += role + ':' + message + seps[i % 2] + '\n'
|
| 190 |
+
else:
|
| 191 |
+
ret += role + ':'
|
| 192 |
+
return ret
|
| 193 |
+
elif self.sep_style == SeparatorStyle.DOLLY:
|
| 194 |
+
seps = [self.sep, self.sep2]
|
| 195 |
+
ret = system_prompt
|
| 196 |
+
for i, (role, message) in enumerate(self.messages):
|
| 197 |
+
if message:
|
| 198 |
+
ret += role + ':\n' + message + seps[i % 2]
|
| 199 |
+
if i % 2 == 1:
|
| 200 |
+
ret += '\n\n'
|
| 201 |
+
else:
|
| 202 |
+
ret += role + ':\n'
|
| 203 |
+
return ret
|
| 204 |
+
elif self.sep_style == SeparatorStyle.PHOENIX:
|
| 205 |
+
ret = system_prompt
|
| 206 |
+
for role, message in self.messages:
|
| 207 |
+
if message:
|
| 208 |
+
ret += role + ': ' + '<s>' + message + '</s>'
|
| 209 |
+
else:
|
| 210 |
+
ret += role + ': ' + '<s>'
|
| 211 |
+
return ret
|
| 212 |
+
elif self.sep_style == SeparatorStyle.ROBIN:
|
| 213 |
+
ret = system_prompt + self.sep
|
| 214 |
+
for role, message in self.messages:
|
| 215 |
+
if message:
|
| 216 |
+
ret += role + ':\n' + message + self.sep
|
| 217 |
+
else:
|
| 218 |
+
ret += role + ':\n'
|
| 219 |
+
return ret
|
| 220 |
+
elif self.sep_style == SeparatorStyle.FALCON_CHAT:
|
| 221 |
+
ret = ''
|
| 222 |
+
if self.system_message:
|
| 223 |
+
ret += system_prompt + self.sep
|
| 224 |
+
for role, message in self.messages:
|
| 225 |
+
if message:
|
| 226 |
+
ret += role + ': ' + message + self.sep
|
| 227 |
+
else:
|
| 228 |
+
ret += role + ':'
|
| 229 |
+
|
| 230 |
+
return ret
|
| 231 |
+
elif self.sep_style == SeparatorStyle.INTERNVL_ZH:
|
| 232 |
+
seps = [self.sep2, self.sep]
|
| 233 |
+
ret = self.system_message + seps[0]
|
| 234 |
+
for i, (role, message) in enumerate(self.messages):
|
| 235 |
+
if message:
|
| 236 |
+
ret += role + ': ' + message + seps[i % 2]
|
| 237 |
+
else:
|
| 238 |
+
ret += role + ':'
|
| 239 |
+
return ret
|
| 240 |
+
elif self.sep_style == SeparatorStyle.MPT:
|
| 241 |
+
ret = system_prompt + self.sep
|
| 242 |
+
for role, message in self.messages:
|
| 243 |
+
if message:
|
| 244 |
+
if type(message) is tuple:
|
| 245 |
+
message, _, _ = message
|
| 246 |
+
ret += role + message + self.sep
|
| 247 |
+
else:
|
| 248 |
+
ret += role
|
| 249 |
+
return ret
|
| 250 |
+
else:
|
| 251 |
+
raise ValueError(f'Invalid style: {self.sep_style}')
|
| 252 |
+
|
| 253 |
+
def set_system_message(self, system_message: str):
|
| 254 |
+
"""Set the system message."""
|
| 255 |
+
self.system_message = system_message
|
| 256 |
+
|
| 257 |
+
def append_message(self, role: str, message: str):
|
| 258 |
+
"""Append a new message."""
|
| 259 |
+
self.messages.append([role, message])
|
| 260 |
+
|
| 261 |
+
def update_last_message(self, message: str):
|
| 262 |
+
"""Update the last output.
|
| 263 |
+
|
| 264 |
+
The last message is typically set to be None when constructing the prompt,
|
| 265 |
+
so we need to update it in-place after getting the response from a model.
|
| 266 |
+
"""
|
| 267 |
+
self.messages[-1][1] = message
|
| 268 |
+
|
| 269 |
+
def to_gradio_chatbot(self):
|
| 270 |
+
"""Convert the conversation to gradio chatbot format."""
|
| 271 |
+
ret = []
|
| 272 |
+
for i, (role, msg) in enumerate(self.messages[self.offset :]):
|
| 273 |
+
if i % 2 == 0:
|
| 274 |
+
ret.append([msg, None])
|
| 275 |
+
else:
|
| 276 |
+
ret[-1][-1] = msg
|
| 277 |
+
return ret
|
| 278 |
+
|
| 279 |
+
def to_openai_api_messages(self):
|
| 280 |
+
"""Convert the conversation to OpenAI chat completion format."""
|
| 281 |
+
ret = [{'role': 'system', 'content': self.system_message}]
|
| 282 |
+
|
| 283 |
+
for i, (_, msg) in enumerate(self.messages[self.offset :]):
|
| 284 |
+
if i % 2 == 0:
|
| 285 |
+
ret.append({'role': 'user', 'content': msg})
|
| 286 |
+
else:
|
| 287 |
+
if msg is not None:
|
| 288 |
+
ret.append({'role': 'assistant', 'content': msg})
|
| 289 |
+
return ret
|
| 290 |
+
|
| 291 |
+
def copy(self):
|
| 292 |
+
return Conversation(
|
| 293 |
+
name=self.name,
|
| 294 |
+
system_template=self.system_template,
|
| 295 |
+
system_message=self.system_message,
|
| 296 |
+
roles=self.roles,
|
| 297 |
+
messages=[[x, y] for x, y in self.messages],
|
| 298 |
+
offset=self.offset,
|
| 299 |
+
sep_style=self.sep_style,
|
| 300 |
+
sep=self.sep,
|
| 301 |
+
sep2=self.sep2,
|
| 302 |
+
stop_str=self.stop_str,
|
| 303 |
+
stop_token_ids=self.stop_token_ids,
|
| 304 |
+
)
|
| 305 |
+
|
| 306 |
+
def dict(self):
|
| 307 |
+
return {
|
| 308 |
+
'template_name': self.name,
|
| 309 |
+
'system_message': self.system_message,
|
| 310 |
+
'roles': self.roles,
|
| 311 |
+
'messages': self.messages,
|
| 312 |
+
'offset': self.offset,
|
| 313 |
+
}
|
| 314 |
+
|
| 315 |
+
|
| 316 |
+
# A global registry for all conversation templates
|
| 317 |
+
conv_templates: Dict[str, Conversation] = {}
|
| 318 |
+
|
| 319 |
+
|
| 320 |
+
def register_conv_template(template: Conversation, override: bool = False):
|
| 321 |
+
"""Register a new conversation template."""
|
| 322 |
+
if not override:
|
| 323 |
+
assert (
|
| 324 |
+
template.name not in conv_templates
|
| 325 |
+
), f'{template.name} has been registered.'
|
| 326 |
+
|
| 327 |
+
conv_templates[template.name] = template
|
| 328 |
+
|
| 329 |
+
|
| 330 |
+
def get_conv_template(name: str) -> Conversation:
|
| 331 |
+
"""Get a conversation template."""
|
| 332 |
+
#print("conv_templates", conv_templates)
|
| 333 |
+
return conv_templates[name].copy()
|
| 334 |
+
|
| 335 |
+
register_conv_template(
|
| 336 |
+
Conversation(
|
| 337 |
+
name='qianfanvl',
|
| 338 |
+
system_template='<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\n{system_message}',
|
| 339 |
+
system_message='你是Qianfan-VL，由百度智能云千帆团队研发的多模态大模型。',
|
| 340 |
+
roles=(
|
| 341 |
+
'<|start_header_id|>user<|end_header_id|>\n\n',
|
| 342 |
+
'<|start_header_id|>assistant<|end_header_id|>\n\n',
|
| 343 |
+
),
|
| 344 |
+
sep_style=SeparatorStyle.MPT,
|
| 345 |
+
sep='<|eot_id|>',
|
| 346 |
+
)
|
| 347 |
+
)
|
| 348 |
+
|
| 349 |
+
register_conv_template(
|
| 350 |
+
Conversation(
|
| 351 |
+
name='qwen',
|
| 352 |
+
system_template='<|im_start|>system<|im_end|>\n\n{system_message}',
|
| 353 |
+
system_message='你是Qianfan-VL，由百度智能云千帆团队研发的多模态大模型。',
|
| 354 |
+
roles=('<|im_start|>user\n', '<|im_start|>assistant\n'),
|
| 355 |
+
sep_style=SeparatorStyle.MPT,
|
| 356 |
+
sep='<|im_end|>',
|
| 357 |
+
)
|
| 358 |
+
)
|
example/scene_ocr.png
ADDED
|
Git LFS Details
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"transformers_version": "4.37.2",
|
| 4 |
+
"repetition_penalty": 1.05,
|
| 5 |
+
"temperature": 0.000001
|
| 6 |
+
}
|
model-00001-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c316574b419d679d54fa2bf2138c4c6ada9b2633214716143b804272b389787c
|
| 3 |
+
size 4980960824
|
model-00002-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f9188a265ff77cdbe2b26ab0d3dc05642a32d2707be4259012e546b9d9774fa
|
| 3 |
+
size 4915920528
|
model-00003-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8b95f8096a848313af81211a24549d69a71c8040820b66959b96baa545d96c8
|
| 3 |
+
size 4947407000
|
model-00004-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6347b44abed47fe0994f0f26c74ee04598101238fe21b4904388106dfdfcbfb7
|
| 3 |
+
size 2772438696
|
model.safetensors.index.json
ADDED
|
@@ -0,0 +1,676 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_size": 17616644096
|
| 4 |
+
},
|
| 5 |
+
"weight_map": {
|
| 6 |
+
"language_model.lm_head.weight": "model-00003-of-00004.safetensors",
|
| 7 |
+
"language_model.model.embed_tokens.weight": "model-00001-of-00004.safetensors",
|
| 8 |
+
"language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 9 |
+
"language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 10 |
+
"language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 11 |
+
"language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 12 |
+
"language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 13 |
+
"language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 14 |
+
"language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 15 |
+
"language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 16 |
+
"language_model.model.layers.0.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 17 |
+
"language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 18 |
+
"language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 19 |
+
"language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 20 |
+
"language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 21 |
+
"language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 22 |
+
"language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 23 |
+
"language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 24 |
+
"language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 25 |
+
"language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 26 |
+
"language_model.model.layers.1.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 27 |
+
"language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 28 |
+
"language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 29 |
+
"language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 30 |
+
"language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 31 |
+
"language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 32 |
+
"language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 33 |
+
"language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 34 |
+
"language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 35 |
+
"language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 36 |
+
"language_model.model.layers.10.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 37 |
+
"language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 38 |
+
"language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 39 |
+
"language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 40 |
+
"language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 41 |
+
"language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 42 |
+
"language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 43 |
+
"language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 44 |
+
"language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 45 |
+
"language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 46 |
+
"language_model.model.layers.11.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 47 |
+
"language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 48 |
+
"language_model.model.layers.12.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 49 |
+
"language_model.model.layers.12.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 50 |
+
"language_model.model.layers.12.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 51 |
+
"language_model.model.layers.12.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 52 |
+
"language_model.model.layers.12.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 53 |
+
"language_model.model.layers.12.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 54 |
+
"language_model.model.layers.12.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 55 |
+
"language_model.model.layers.12.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 56 |
+
"language_model.model.layers.12.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 57 |
+
"language_model.model.layers.12.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 58 |
+
"language_model.model.layers.13.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 59 |
+
"language_model.model.layers.13.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 60 |
+
"language_model.model.layers.13.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 61 |
+
"language_model.model.layers.13.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 62 |
+
"language_model.model.layers.13.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 63 |
+
"language_model.model.layers.13.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 64 |
+
"language_model.model.layers.13.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 65 |
+
"language_model.model.layers.13.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 66 |
+
"language_model.model.layers.13.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 67 |
+
"language_model.model.layers.13.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 68 |
+
"language_model.model.layers.14.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 69 |
+
"language_model.model.layers.14.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 70 |
+
"language_model.model.layers.14.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 71 |
+
"language_model.model.layers.14.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 72 |
+
"language_model.model.layers.14.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 73 |
+
"language_model.model.layers.14.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 74 |
+
"language_model.model.layers.14.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 75 |
+
"language_model.model.layers.14.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 76 |
+
"language_model.model.layers.14.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 77 |
+
"language_model.model.layers.14.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 78 |
+
"language_model.model.layers.15.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 79 |
+
"language_model.model.layers.15.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 80 |
+
"language_model.model.layers.15.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 81 |
+
"language_model.model.layers.15.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 82 |
+
"language_model.model.layers.15.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 83 |
+
"language_model.model.layers.15.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 84 |
+
"language_model.model.layers.15.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 85 |
+
"language_model.model.layers.15.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 86 |
+
"language_model.model.layers.15.self_attn.rotary_emb.inv_freq": "model-00001-of-00004.safetensors",
|
| 87 |
+
"language_model.model.layers.15.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 88 |
+
"language_model.model.layers.16.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 89 |
+
"language_model.model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 90 |
+
"language_model.model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 91 |
+
"language_model.model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 92 |
+
"language_model.model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 93 |
+
"language_model.model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 94 |
+
"language_model.model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 95 |
+
"language_model.model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 96 |
+
"language_model.model.layers.16.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 97 |
+
"language_model.model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 98 |
+
"language_model.model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 99 |
+
"language_model.model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 100 |
+
"language_model.model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 101 |
+
"language_model.model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 102 |
+
"language_model.model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 103 |
+
"language_model.model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 104 |
+
"language_model.model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 105 |
+
"language_model.model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 106 |
+
"language_model.model.layers.17.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 107 |
+
"language_model.model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 108 |
+
"language_model.model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 109 |
+
"language_model.model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 110 |
+
"language_model.model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 111 |
+
"language_model.model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 112 |
+
"language_model.model.layers.18.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 113 |
+
"language_model.model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 114 |
+
"language_model.model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 115 |
+
"language_model.model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 116 |
+
"language_model.model.layers.18.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 117 |
+
"language_model.model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 118 |
+
"language_model.model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 119 |
+
"language_model.model.layers.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 120 |
+
"language_model.model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 121 |
+
"language_model.model.layers.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 122 |
+
"language_model.model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 123 |
+
"language_model.model.layers.19.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 124 |
+
"language_model.model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 125 |
+
"language_model.model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 126 |
+
"language_model.model.layers.19.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 127 |
+
"language_model.model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 128 |
+
"language_model.model.layers.2.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 129 |
+
"language_model.model.layers.2.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 130 |
+
"language_model.model.layers.2.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 131 |
+
"language_model.model.layers.2.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 132 |
+
"language_model.model.layers.2.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 133 |
+
"language_model.model.layers.2.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 134 |
+
"language_model.model.layers.2.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 135 |
+
"language_model.model.layers.2.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 136 |
+
"language_model.model.layers.2.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 137 |
+
"language_model.model.layers.2.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 138 |
+
"language_model.model.layers.20.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 139 |
+
"language_model.model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 140 |
+
"language_model.model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 141 |
+
"language_model.model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 142 |
+
"language_model.model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 143 |
+
"language_model.model.layers.20.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 144 |
+
"language_model.model.layers.20.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 145 |
+
"language_model.model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 146 |
+
"language_model.model.layers.20.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 147 |
+
"language_model.model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 148 |
+
"language_model.model.layers.21.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 149 |
+
"language_model.model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 150 |
+
"language_model.model.layers.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 151 |
+
"language_model.model.layers.21.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 152 |
+
"language_model.model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 153 |
+
"language_model.model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 154 |
+
"language_model.model.layers.21.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 155 |
+
"language_model.model.layers.21.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 156 |
+
"language_model.model.layers.21.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 157 |
+
"language_model.model.layers.21.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 158 |
+
"language_model.model.layers.22.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 159 |
+
"language_model.model.layers.22.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 160 |
+
"language_model.model.layers.22.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 161 |
+
"language_model.model.layers.22.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 162 |
+
"language_model.model.layers.22.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 163 |
+
"language_model.model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 164 |
+
"language_model.model.layers.22.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 165 |
+
"language_model.model.layers.22.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 166 |
+
"language_model.model.layers.22.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 167 |
+
"language_model.model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 168 |
+
"language_model.model.layers.23.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 169 |
+
"language_model.model.layers.23.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 170 |
+
"language_model.model.layers.23.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 171 |
+
"language_model.model.layers.23.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 172 |
+
"language_model.model.layers.23.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 173 |
+
"language_model.model.layers.23.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 174 |
+
"language_model.model.layers.23.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 175 |
+
"language_model.model.layers.23.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 176 |
+
"language_model.model.layers.23.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 177 |
+
"language_model.model.layers.23.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 178 |
+
"language_model.model.layers.24.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 179 |
+
"language_model.model.layers.24.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 180 |
+
"language_model.model.layers.24.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 181 |
+
"language_model.model.layers.24.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 182 |
+
"language_model.model.layers.24.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 183 |
+
"language_model.model.layers.24.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 184 |
+
"language_model.model.layers.24.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 185 |
+
"language_model.model.layers.24.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 186 |
+
"language_model.model.layers.24.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 187 |
+
"language_model.model.layers.24.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 188 |
+
"language_model.model.layers.25.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 189 |
+
"language_model.model.layers.25.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 190 |
+
"language_model.model.layers.25.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 191 |
+
"language_model.model.layers.25.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 192 |
+
"language_model.model.layers.25.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 193 |
+
"language_model.model.layers.25.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 194 |
+
"language_model.model.layers.25.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 195 |
+
"language_model.model.layers.25.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 196 |
+
"language_model.model.layers.25.self_attn.rotary_emb.inv_freq": "model-00002-of-00004.safetensors",
|
| 197 |
+
"language_model.model.layers.25.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 198 |
+
"language_model.model.layers.26.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 199 |
+
"language_model.model.layers.26.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 200 |
+
"language_model.model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 201 |
+
"language_model.model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 202 |
+
"language_model.model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 203 |
+
"language_model.model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 204 |
+
"language_model.model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 205 |
+
"language_model.model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 206 |
+
"language_model.model.layers.26.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 207 |
+
"language_model.model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 208 |
+
"language_model.model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 209 |
+
"language_model.model.layers.27.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 210 |
+
"language_model.model.layers.27.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 211 |
+
"language_model.model.layers.27.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 212 |
+
"language_model.model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 213 |
+
"language_model.model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 214 |
+
"language_model.model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 215 |
+
"language_model.model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 216 |
+
"language_model.model.layers.27.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 217 |
+
"language_model.model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 218 |
+
"language_model.model.layers.28.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 219 |
+
"language_model.model.layers.28.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 220 |
+
"language_model.model.layers.28.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 221 |
+
"language_model.model.layers.28.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 222 |
+
"language_model.model.layers.28.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 223 |
+
"language_model.model.layers.28.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
|
| 224 |
+
"language_model.model.layers.28.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
|
| 225 |
+
"language_model.model.layers.28.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
|
| 226 |
+
"language_model.model.layers.28.self_attn.rotary_emb.inv_freq": "model-00004-of-00004.safetensors",
|
| 227 |
+
"language_model.model.layers.28.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
|
| 228 |
+
"language_model.model.layers.29.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 229 |
+
"language_model.model.layers.29.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 230 |
+
"language_model.model.layers.29.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 231 |
+
"language_model.model.layers.29.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 232 |
+
"language_model.model.layers.29.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 233 |
+
"language_model.model.layers.29.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
|
| 234 |
+
"language_model.model.layers.29.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
|
| 235 |
+
"language_model.model.layers.29.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
|
| 236 |
+
"language_model.model.layers.29.self_attn.rotary_emb.inv_freq": "model-00004-of-00004.safetensors",
|
| 237 |
+
"language_model.model.layers.29.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
|
| 238 |
+
"language_model.model.layers.3.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 239 |
+
"language_model.model.layers.3.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 240 |
+
"language_model.model.layers.3.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 241 |
+
"language_model.model.layers.3.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 242 |
+
"language_model.model.layers.3.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 243 |
+
"language_model.model.layers.3.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 244 |
+
"language_model.model.layers.3.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 245 |
+
"language_model.model.layers.3.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 246 |
+
"language_model.model.layers.3.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 247 |
+
"language_model.model.layers.3.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 248 |
+
"language_model.model.layers.30.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 249 |
+
"language_model.model.layers.30.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 250 |
+
"language_model.model.layers.30.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 251 |
+
"language_model.model.layers.30.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 252 |
+
"language_model.model.layers.30.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 253 |
+
"language_model.model.layers.30.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
|
| 254 |
+
"language_model.model.layers.30.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
|
| 255 |
+
"language_model.model.layers.30.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
|
| 256 |
+
"language_model.model.layers.30.self_attn.rotary_emb.inv_freq": "model-00004-of-00004.safetensors",
|
| 257 |
+
"language_model.model.layers.30.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
|
| 258 |
+
"language_model.model.layers.31.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 259 |
+
"language_model.model.layers.31.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 260 |
+
"language_model.model.layers.31.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 261 |
+
"language_model.model.layers.31.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 262 |
+
"language_model.model.layers.31.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 263 |
+
"language_model.model.layers.31.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
|
| 264 |
+
"language_model.model.layers.31.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
|
| 265 |
+
"language_model.model.layers.31.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
|
| 266 |
+
"language_model.model.layers.31.self_attn.rotary_emb.inv_freq": "model-00004-of-00004.safetensors",
|
| 267 |
+
"language_model.model.layers.31.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
|
| 268 |
+
"language_model.model.layers.4.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 269 |
+
"language_model.model.layers.4.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 270 |
+
"language_model.model.layers.4.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 271 |
+
"language_model.model.layers.4.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 272 |
+
"language_model.model.layers.4.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 273 |
+
"language_model.model.layers.4.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 274 |
+
"language_model.model.layers.4.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 275 |
+
"language_model.model.layers.4.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 276 |
+
"language_model.model.layers.4.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 277 |
+
"language_model.model.layers.4.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 278 |
+
"language_model.model.layers.5.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 279 |
+
"language_model.model.layers.5.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 280 |
+
"language_model.model.layers.5.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 281 |
+
"language_model.model.layers.5.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 282 |
+
"language_model.model.layers.5.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 283 |
+
"language_model.model.layers.5.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 284 |
+
"language_model.model.layers.5.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 285 |
+
"language_model.model.layers.5.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 286 |
+
"language_model.model.layers.5.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 287 |
+
"language_model.model.layers.5.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 288 |
+
"language_model.model.layers.6.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 289 |
+
"language_model.model.layers.6.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 290 |
+
"language_model.model.layers.6.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 291 |
+
"language_model.model.layers.6.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 292 |
+
"language_model.model.layers.6.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 293 |
+
"language_model.model.layers.6.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 294 |
+
"language_model.model.layers.6.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 295 |
+
"language_model.model.layers.6.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 296 |
+
"language_model.model.layers.6.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 297 |
+
"language_model.model.layers.6.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 298 |
+
"language_model.model.layers.7.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 299 |
+
"language_model.model.layers.7.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 300 |
+
"language_model.model.layers.7.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 301 |
+
"language_model.model.layers.7.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 302 |
+
"language_model.model.layers.7.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 303 |
+
"language_model.model.layers.7.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 304 |
+
"language_model.model.layers.7.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 305 |
+
"language_model.model.layers.7.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 306 |
+
"language_model.model.layers.7.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 307 |
+
"language_model.model.layers.7.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 308 |
+
"language_model.model.layers.8.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 309 |
+
"language_model.model.layers.8.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 310 |
+
"language_model.model.layers.8.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 311 |
+
"language_model.model.layers.8.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 312 |
+
"language_model.model.layers.8.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 313 |
+
"language_model.model.layers.8.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 314 |
+
"language_model.model.layers.8.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 315 |
+
"language_model.model.layers.8.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 316 |
+
"language_model.model.layers.8.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 317 |
+
"language_model.model.layers.8.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 318 |
+
"language_model.model.layers.9.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 319 |
+
"language_model.model.layers.9.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 320 |
+
"language_model.model.layers.9.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 321 |
+
"language_model.model.layers.9.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 322 |
+
"language_model.model.layers.9.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 323 |
+
"language_model.model.layers.9.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 324 |
+
"language_model.model.layers.9.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 325 |
+
"language_model.model.layers.9.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 326 |
+
"language_model.model.layers.9.self_attn.rotary_emb.inv_freq": "model-00003-of-00004.safetensors",
|
| 327 |
+
"language_model.model.layers.9.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 328 |
+
"language_model.model.norm.weight": "model-00004-of-00004.safetensors",
|
| 329 |
+
"mlp1.0.bias": "model-00004-of-00004.safetensors",
|
| 330 |
+
"mlp1.0.weight": "model-00004-of-00004.safetensors",
|
| 331 |
+
"mlp1.1.bias": "model-00004-of-00004.safetensors",
|
| 332 |
+
"mlp1.1.weight": "model-00004-of-00004.safetensors",
|
| 333 |
+
"mlp1.3.bias": "model-00004-of-00004.safetensors",
|
| 334 |
+
"mlp1.3.weight": "model-00004-of-00004.safetensors",
|
| 335 |
+
"vision_model.embeddings.class_embedding": "model-00004-of-00004.safetensors",
|
| 336 |
+
"vision_model.embeddings.patch_embedding.bias": "model-00004-of-00004.safetensors",
|
| 337 |
+
"vision_model.embeddings.patch_embedding.weight": "model-00004-of-00004.safetensors",
|
| 338 |
+
"vision_model.embeddings.position_embedding": "model-00004-of-00004.safetensors",
|
| 339 |
+
"vision_model.encoder.layers.0.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 340 |
+
"vision_model.encoder.layers.0.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 341 |
+
"vision_model.encoder.layers.0.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 342 |
+
"vision_model.encoder.layers.0.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 343 |
+
"vision_model.encoder.layers.0.ls1": "model-00004-of-00004.safetensors",
|
| 344 |
+
"vision_model.encoder.layers.0.ls2": "model-00004-of-00004.safetensors",
|
| 345 |
+
"vision_model.encoder.layers.0.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 346 |
+
"vision_model.encoder.layers.0.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 347 |
+
"vision_model.encoder.layers.0.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 348 |
+
"vision_model.encoder.layers.0.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 349 |
+
"vision_model.encoder.layers.0.norm1.bias": "model-00004-of-00004.safetensors",
|
| 350 |
+
"vision_model.encoder.layers.0.norm1.weight": "model-00004-of-00004.safetensors",
|
| 351 |
+
"vision_model.encoder.layers.0.norm2.bias": "model-00004-of-00004.safetensors",
|
| 352 |
+
"vision_model.encoder.layers.0.norm2.weight": "model-00004-of-00004.safetensors",
|
| 353 |
+
"vision_model.encoder.layers.1.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 354 |
+
"vision_model.encoder.layers.1.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 355 |
+
"vision_model.encoder.layers.1.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 356 |
+
"vision_model.encoder.layers.1.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 357 |
+
"vision_model.encoder.layers.1.ls1": "model-00004-of-00004.safetensors",
|
| 358 |
+
"vision_model.encoder.layers.1.ls2": "model-00004-of-00004.safetensors",
|
| 359 |
+
"vision_model.encoder.layers.1.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 360 |
+
"vision_model.encoder.layers.1.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 361 |
+
"vision_model.encoder.layers.1.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 362 |
+
"vision_model.encoder.layers.1.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 363 |
+
"vision_model.encoder.layers.1.norm1.bias": "model-00004-of-00004.safetensors",
|
| 364 |
+
"vision_model.encoder.layers.1.norm1.weight": "model-00004-of-00004.safetensors",
|
| 365 |
+
"vision_model.encoder.layers.1.norm2.bias": "model-00004-of-00004.safetensors",
|
| 366 |
+
"vision_model.encoder.layers.1.norm2.weight": "model-00004-of-00004.safetensors",
|
| 367 |
+
"vision_model.encoder.layers.10.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 368 |
+
"vision_model.encoder.layers.10.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 369 |
+
"vision_model.encoder.layers.10.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 370 |
+
"vision_model.encoder.layers.10.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 371 |
+
"vision_model.encoder.layers.10.ls1": "model-00004-of-00004.safetensors",
|
| 372 |
+
"vision_model.encoder.layers.10.ls2": "model-00004-of-00004.safetensors",
|
| 373 |
+
"vision_model.encoder.layers.10.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 374 |
+
"vision_model.encoder.layers.10.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 375 |
+
"vision_model.encoder.layers.10.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 376 |
+
"vision_model.encoder.layers.10.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 377 |
+
"vision_model.encoder.layers.10.norm1.bias": "model-00004-of-00004.safetensors",
|
| 378 |
+
"vision_model.encoder.layers.10.norm1.weight": "model-00004-of-00004.safetensors",
|
| 379 |
+
"vision_model.encoder.layers.10.norm2.bias": "model-00004-of-00004.safetensors",
|
| 380 |
+
"vision_model.encoder.layers.10.norm2.weight": "model-00004-of-00004.safetensors",
|
| 381 |
+
"vision_model.encoder.layers.11.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 382 |
+
"vision_model.encoder.layers.11.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 383 |
+
"vision_model.encoder.layers.11.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 384 |
+
"vision_model.encoder.layers.11.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 385 |
+
"vision_model.encoder.layers.11.ls1": "model-00004-of-00004.safetensors",
|
| 386 |
+
"vision_model.encoder.layers.11.ls2": "model-00004-of-00004.safetensors",
|
| 387 |
+
"vision_model.encoder.layers.11.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 388 |
+
"vision_model.encoder.layers.11.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 389 |
+
"vision_model.encoder.layers.11.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 390 |
+
"vision_model.encoder.layers.11.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 391 |
+
"vision_model.encoder.layers.11.norm1.bias": "model-00004-of-00004.safetensors",
|
| 392 |
+
"vision_model.encoder.layers.11.norm1.weight": "model-00004-of-00004.safetensors",
|
| 393 |
+
"vision_model.encoder.layers.11.norm2.bias": "model-00004-of-00004.safetensors",
|
| 394 |
+
"vision_model.encoder.layers.11.norm2.weight": "model-00004-of-00004.safetensors",
|
| 395 |
+
"vision_model.encoder.layers.12.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 396 |
+
"vision_model.encoder.layers.12.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 397 |
+
"vision_model.encoder.layers.12.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 398 |
+
"vision_model.encoder.layers.12.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 399 |
+
"vision_model.encoder.layers.12.ls1": "model-00004-of-00004.safetensors",
|
| 400 |
+
"vision_model.encoder.layers.12.ls2": "model-00004-of-00004.safetensors",
|
| 401 |
+
"vision_model.encoder.layers.12.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 402 |
+
"vision_model.encoder.layers.12.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 403 |
+
"vision_model.encoder.layers.12.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 404 |
+
"vision_model.encoder.layers.12.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 405 |
+
"vision_model.encoder.layers.12.norm1.bias": "model-00004-of-00004.safetensors",
|
| 406 |
+
"vision_model.encoder.layers.12.norm1.weight": "model-00004-of-00004.safetensors",
|
| 407 |
+
"vision_model.encoder.layers.12.norm2.bias": "model-00004-of-00004.safetensors",
|
| 408 |
+
"vision_model.encoder.layers.12.norm2.weight": "model-00004-of-00004.safetensors",
|
| 409 |
+
"vision_model.encoder.layers.13.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 410 |
+
"vision_model.encoder.layers.13.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 411 |
+
"vision_model.encoder.layers.13.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 412 |
+
"vision_model.encoder.layers.13.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 413 |
+
"vision_model.encoder.layers.13.ls1": "model-00004-of-00004.safetensors",
|
| 414 |
+
"vision_model.encoder.layers.13.ls2": "model-00004-of-00004.safetensors",
|
| 415 |
+
"vision_model.encoder.layers.13.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 416 |
+
"vision_model.encoder.layers.13.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 417 |
+
"vision_model.encoder.layers.13.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 418 |
+
"vision_model.encoder.layers.13.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 419 |
+
"vision_model.encoder.layers.13.norm1.bias": "model-00004-of-00004.safetensors",
|
| 420 |
+
"vision_model.encoder.layers.13.norm1.weight": "model-00004-of-00004.safetensors",
|
| 421 |
+
"vision_model.encoder.layers.13.norm2.bias": "model-00004-of-00004.safetensors",
|
| 422 |
+
"vision_model.encoder.layers.13.norm2.weight": "model-00004-of-00004.safetensors",
|
| 423 |
+
"vision_model.encoder.layers.14.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 424 |
+
"vision_model.encoder.layers.14.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 425 |
+
"vision_model.encoder.layers.14.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 426 |
+
"vision_model.encoder.layers.14.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 427 |
+
"vision_model.encoder.layers.14.ls1": "model-00004-of-00004.safetensors",
|
| 428 |
+
"vision_model.encoder.layers.14.ls2": "model-00004-of-00004.safetensors",
|
| 429 |
+
"vision_model.encoder.layers.14.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 430 |
+
"vision_model.encoder.layers.14.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 431 |
+
"vision_model.encoder.layers.14.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 432 |
+
"vision_model.encoder.layers.14.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 433 |
+
"vision_model.encoder.layers.14.norm1.bias": "model-00004-of-00004.safetensors",
|
| 434 |
+
"vision_model.encoder.layers.14.norm1.weight": "model-00004-of-00004.safetensors",
|
| 435 |
+
"vision_model.encoder.layers.14.norm2.bias": "model-00004-of-00004.safetensors",
|
| 436 |
+
"vision_model.encoder.layers.14.norm2.weight": "model-00004-of-00004.safetensors",
|
| 437 |
+
"vision_model.encoder.layers.15.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 438 |
+
"vision_model.encoder.layers.15.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 439 |
+
"vision_model.encoder.layers.15.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 440 |
+
"vision_model.encoder.layers.15.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 441 |
+
"vision_model.encoder.layers.15.ls1": "model-00004-of-00004.safetensors",
|
| 442 |
+
"vision_model.encoder.layers.15.ls2": "model-00004-of-00004.safetensors",
|
| 443 |
+
"vision_model.encoder.layers.15.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 444 |
+
"vision_model.encoder.layers.15.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 445 |
+
"vision_model.encoder.layers.15.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 446 |
+
"vision_model.encoder.layers.15.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 447 |
+
"vision_model.encoder.layers.15.norm1.bias": "model-00004-of-00004.safetensors",
|
| 448 |
+
"vision_model.encoder.layers.15.norm1.weight": "model-00004-of-00004.safetensors",
|
| 449 |
+
"vision_model.encoder.layers.15.norm2.bias": "model-00004-of-00004.safetensors",
|
| 450 |
+
"vision_model.encoder.layers.15.norm2.weight": "model-00004-of-00004.safetensors",
|
| 451 |
+
"vision_model.encoder.layers.16.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 452 |
+
"vision_model.encoder.layers.16.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 453 |
+
"vision_model.encoder.layers.16.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 454 |
+
"vision_model.encoder.layers.16.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 455 |
+
"vision_model.encoder.layers.16.ls1": "model-00004-of-00004.safetensors",
|
| 456 |
+
"vision_model.encoder.layers.16.ls2": "model-00004-of-00004.safetensors",
|
| 457 |
+
"vision_model.encoder.layers.16.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 458 |
+
"vision_model.encoder.layers.16.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 459 |
+
"vision_model.encoder.layers.16.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 460 |
+
"vision_model.encoder.layers.16.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 461 |
+
"vision_model.encoder.layers.16.norm1.bias": "model-00004-of-00004.safetensors",
|
| 462 |
+
"vision_model.encoder.layers.16.norm1.weight": "model-00004-of-00004.safetensors",
|
| 463 |
+
"vision_model.encoder.layers.16.norm2.bias": "model-00004-of-00004.safetensors",
|
| 464 |
+
"vision_model.encoder.layers.16.norm2.weight": "model-00004-of-00004.safetensors",
|
| 465 |
+
"vision_model.encoder.layers.17.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 466 |
+
"vision_model.encoder.layers.17.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 467 |
+
"vision_model.encoder.layers.17.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 468 |
+
"vision_model.encoder.layers.17.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 469 |
+
"vision_model.encoder.layers.17.ls1": "model-00004-of-00004.safetensors",
|
| 470 |
+
"vision_model.encoder.layers.17.ls2": "model-00004-of-00004.safetensors",
|
| 471 |
+
"vision_model.encoder.layers.17.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 472 |
+
"vision_model.encoder.layers.17.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 473 |
+
"vision_model.encoder.layers.17.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 474 |
+
"vision_model.encoder.layers.17.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 475 |
+
"vision_model.encoder.layers.17.norm1.bias": "model-00004-of-00004.safetensors",
|
| 476 |
+
"vision_model.encoder.layers.17.norm1.weight": "model-00004-of-00004.safetensors",
|
| 477 |
+
"vision_model.encoder.layers.17.norm2.bias": "model-00004-of-00004.safetensors",
|
| 478 |
+
"vision_model.encoder.layers.17.norm2.weight": "model-00004-of-00004.safetensors",
|
| 479 |
+
"vision_model.encoder.layers.18.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 480 |
+
"vision_model.encoder.layers.18.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 481 |
+
"vision_model.encoder.layers.18.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 482 |
+
"vision_model.encoder.layers.18.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 483 |
+
"vision_model.encoder.layers.18.ls1": "model-00004-of-00004.safetensors",
|
| 484 |
+
"vision_model.encoder.layers.18.ls2": "model-00004-of-00004.safetensors",
|
| 485 |
+
"vision_model.encoder.layers.18.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 486 |
+
"vision_model.encoder.layers.18.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 487 |
+
"vision_model.encoder.layers.18.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 488 |
+
"vision_model.encoder.layers.18.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 489 |
+
"vision_model.encoder.layers.18.norm1.bias": "model-00004-of-00004.safetensors",
|
| 490 |
+
"vision_model.encoder.layers.18.norm1.weight": "model-00004-of-00004.safetensors",
|
| 491 |
+
"vision_model.encoder.layers.18.norm2.bias": "model-00004-of-00004.safetensors",
|
| 492 |
+
"vision_model.encoder.layers.18.norm2.weight": "model-00004-of-00004.safetensors",
|
| 493 |
+
"vision_model.encoder.layers.19.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 494 |
+
"vision_model.encoder.layers.19.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 495 |
+
"vision_model.encoder.layers.19.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 496 |
+
"vision_model.encoder.layers.19.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 497 |
+
"vision_model.encoder.layers.19.ls1": "model-00004-of-00004.safetensors",
|
| 498 |
+
"vision_model.encoder.layers.19.ls2": "model-00004-of-00004.safetensors",
|
| 499 |
+
"vision_model.encoder.layers.19.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 500 |
+
"vision_model.encoder.layers.19.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 501 |
+
"vision_model.encoder.layers.19.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 502 |
+
"vision_model.encoder.layers.19.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 503 |
+
"vision_model.encoder.layers.19.norm1.bias": "model-00004-of-00004.safetensors",
|
| 504 |
+
"vision_model.encoder.layers.19.norm1.weight": "model-00004-of-00004.safetensors",
|
| 505 |
+
"vision_model.encoder.layers.19.norm2.bias": "model-00004-of-00004.safetensors",
|
| 506 |
+
"vision_model.encoder.layers.19.norm2.weight": "model-00004-of-00004.safetensors",
|
| 507 |
+
"vision_model.encoder.layers.2.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 508 |
+
"vision_model.encoder.layers.2.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 509 |
+
"vision_model.encoder.layers.2.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 510 |
+
"vision_model.encoder.layers.2.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 511 |
+
"vision_model.encoder.layers.2.ls1": "model-00004-of-00004.safetensors",
|
| 512 |
+
"vision_model.encoder.layers.2.ls2": "model-00004-of-00004.safetensors",
|
| 513 |
+
"vision_model.encoder.layers.2.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 514 |
+
"vision_model.encoder.layers.2.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 515 |
+
"vision_model.encoder.layers.2.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 516 |
+
"vision_model.encoder.layers.2.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 517 |
+
"vision_model.encoder.layers.2.norm1.bias": "model-00004-of-00004.safetensors",
|
| 518 |
+
"vision_model.encoder.layers.2.norm1.weight": "model-00004-of-00004.safetensors",
|
| 519 |
+
"vision_model.encoder.layers.2.norm2.bias": "model-00004-of-00004.safetensors",
|
| 520 |
+
"vision_model.encoder.layers.2.norm2.weight": "model-00004-of-00004.safetensors",
|
| 521 |
+
"vision_model.encoder.layers.20.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 522 |
+
"vision_model.encoder.layers.20.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 523 |
+
"vision_model.encoder.layers.20.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 524 |
+
"vision_model.encoder.layers.20.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 525 |
+
"vision_model.encoder.layers.20.ls1": "model-00004-of-00004.safetensors",
|
| 526 |
+
"vision_model.encoder.layers.20.ls2": "model-00004-of-00004.safetensors",
|
| 527 |
+
"vision_model.encoder.layers.20.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 528 |
+
"vision_model.encoder.layers.20.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 529 |
+
"vision_model.encoder.layers.20.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 530 |
+
"vision_model.encoder.layers.20.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 531 |
+
"vision_model.encoder.layers.20.norm1.bias": "model-00004-of-00004.safetensors",
|
| 532 |
+
"vision_model.encoder.layers.20.norm1.weight": "model-00004-of-00004.safetensors",
|
| 533 |
+
"vision_model.encoder.layers.20.norm2.bias": "model-00004-of-00004.safetensors",
|
| 534 |
+
"vision_model.encoder.layers.20.norm2.weight": "model-00004-of-00004.safetensors",
|
| 535 |
+
"vision_model.encoder.layers.21.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 536 |
+
"vision_model.encoder.layers.21.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 537 |
+
"vision_model.encoder.layers.21.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 538 |
+
"vision_model.encoder.layers.21.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 539 |
+
"vision_model.encoder.layers.21.ls1": "model-00004-of-00004.safetensors",
|
| 540 |
+
"vision_model.encoder.layers.21.ls2": "model-00004-of-00004.safetensors",
|
| 541 |
+
"vision_model.encoder.layers.21.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 542 |
+
"vision_model.encoder.layers.21.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 543 |
+
"vision_model.encoder.layers.21.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 544 |
+
"vision_model.encoder.layers.21.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 545 |
+
"vision_model.encoder.layers.21.norm1.bias": "model-00004-of-00004.safetensors",
|
| 546 |
+
"vision_model.encoder.layers.21.norm1.weight": "model-00004-of-00004.safetensors",
|
| 547 |
+
"vision_model.encoder.layers.21.norm2.bias": "model-00004-of-00004.safetensors",
|
| 548 |
+
"vision_model.encoder.layers.21.norm2.weight": "model-00004-of-00004.safetensors",
|
| 549 |
+
"vision_model.encoder.layers.22.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 550 |
+
"vision_model.encoder.layers.22.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 551 |
+
"vision_model.encoder.layers.22.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 552 |
+
"vision_model.encoder.layers.22.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 553 |
+
"vision_model.encoder.layers.22.ls1": "model-00004-of-00004.safetensors",
|
| 554 |
+
"vision_model.encoder.layers.22.ls2": "model-00004-of-00004.safetensors",
|
| 555 |
+
"vision_model.encoder.layers.22.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 556 |
+
"vision_model.encoder.layers.22.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 557 |
+
"vision_model.encoder.layers.22.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 558 |
+
"vision_model.encoder.layers.22.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 559 |
+
"vision_model.encoder.layers.22.norm1.bias": "model-00004-of-00004.safetensors",
|
| 560 |
+
"vision_model.encoder.layers.22.norm1.weight": "model-00004-of-00004.safetensors",
|
| 561 |
+
"vision_model.encoder.layers.22.norm2.bias": "model-00004-of-00004.safetensors",
|
| 562 |
+
"vision_model.encoder.layers.22.norm2.weight": "model-00004-of-00004.safetensors",
|
| 563 |
+
"vision_model.encoder.layers.23.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 564 |
+
"vision_model.encoder.layers.23.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 565 |
+
"vision_model.encoder.layers.23.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 566 |
+
"vision_model.encoder.layers.23.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 567 |
+
"vision_model.encoder.layers.23.ls1": "model-00004-of-00004.safetensors",
|
| 568 |
+
"vision_model.encoder.layers.23.ls2": "model-00004-of-00004.safetensors",
|
| 569 |
+
"vision_model.encoder.layers.23.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 570 |
+
"vision_model.encoder.layers.23.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 571 |
+
"vision_model.encoder.layers.23.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 572 |
+
"vision_model.encoder.layers.23.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 573 |
+
"vision_model.encoder.layers.23.norm1.bias": "model-00004-of-00004.safetensors",
|
| 574 |
+
"vision_model.encoder.layers.23.norm1.weight": "model-00004-of-00004.safetensors",
|
| 575 |
+
"vision_model.encoder.layers.23.norm2.bias": "model-00004-of-00004.safetensors",
|
| 576 |
+
"vision_model.encoder.layers.23.norm2.weight": "model-00004-of-00004.safetensors",
|
| 577 |
+
"vision_model.encoder.layers.3.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 578 |
+
"vision_model.encoder.layers.3.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 579 |
+
"vision_model.encoder.layers.3.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 580 |
+
"vision_model.encoder.layers.3.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 581 |
+
"vision_model.encoder.layers.3.ls1": "model-00004-of-00004.safetensors",
|
| 582 |
+
"vision_model.encoder.layers.3.ls2": "model-00004-of-00004.safetensors",
|
| 583 |
+
"vision_model.encoder.layers.3.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 584 |
+
"vision_model.encoder.layers.3.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 585 |
+
"vision_model.encoder.layers.3.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 586 |
+
"vision_model.encoder.layers.3.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 587 |
+
"vision_model.encoder.layers.3.norm1.bias": "model-00004-of-00004.safetensors",
|
| 588 |
+
"vision_model.encoder.layers.3.norm1.weight": "model-00004-of-00004.safetensors",
|
| 589 |
+
"vision_model.encoder.layers.3.norm2.bias": "model-00004-of-00004.safetensors",
|
| 590 |
+
"vision_model.encoder.layers.3.norm2.weight": "model-00004-of-00004.safetensors",
|
| 591 |
+
"vision_model.encoder.layers.4.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 592 |
+
"vision_model.encoder.layers.4.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 593 |
+
"vision_model.encoder.layers.4.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 594 |
+
"vision_model.encoder.layers.4.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 595 |
+
"vision_model.encoder.layers.4.ls1": "model-00004-of-00004.safetensors",
|
| 596 |
+
"vision_model.encoder.layers.4.ls2": "model-00004-of-00004.safetensors",
|
| 597 |
+
"vision_model.encoder.layers.4.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 598 |
+
"vision_model.encoder.layers.4.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 599 |
+
"vision_model.encoder.layers.4.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 600 |
+
"vision_model.encoder.layers.4.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 601 |
+
"vision_model.encoder.layers.4.norm1.bias": "model-00004-of-00004.safetensors",
|
| 602 |
+
"vision_model.encoder.layers.4.norm1.weight": "model-00004-of-00004.safetensors",
|
| 603 |
+
"vision_model.encoder.layers.4.norm2.bias": "model-00004-of-00004.safetensors",
|
| 604 |
+
"vision_model.encoder.layers.4.norm2.weight": "model-00004-of-00004.safetensors",
|
| 605 |
+
"vision_model.encoder.layers.5.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 606 |
+
"vision_model.encoder.layers.5.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 607 |
+
"vision_model.encoder.layers.5.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 608 |
+
"vision_model.encoder.layers.5.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 609 |
+
"vision_model.encoder.layers.5.ls1": "model-00004-of-00004.safetensors",
|
| 610 |
+
"vision_model.encoder.layers.5.ls2": "model-00004-of-00004.safetensors",
|
| 611 |
+
"vision_model.encoder.layers.5.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 612 |
+
"vision_model.encoder.layers.5.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 613 |
+
"vision_model.encoder.layers.5.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 614 |
+
"vision_model.encoder.layers.5.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 615 |
+
"vision_model.encoder.layers.5.norm1.bias": "model-00004-of-00004.safetensors",
|
| 616 |
+
"vision_model.encoder.layers.5.norm1.weight": "model-00004-of-00004.safetensors",
|
| 617 |
+
"vision_model.encoder.layers.5.norm2.bias": "model-00004-of-00004.safetensors",
|
| 618 |
+
"vision_model.encoder.layers.5.norm2.weight": "model-00004-of-00004.safetensors",
|
| 619 |
+
"vision_model.encoder.layers.6.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 620 |
+
"vision_model.encoder.layers.6.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 621 |
+
"vision_model.encoder.layers.6.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 622 |
+
"vision_model.encoder.layers.6.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 623 |
+
"vision_model.encoder.layers.6.ls1": "model-00004-of-00004.safetensors",
|
| 624 |
+
"vision_model.encoder.layers.6.ls2": "model-00004-of-00004.safetensors",
|
| 625 |
+
"vision_model.encoder.layers.6.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 626 |
+
"vision_model.encoder.layers.6.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 627 |
+
"vision_model.encoder.layers.6.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 628 |
+
"vision_model.encoder.layers.6.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 629 |
+
"vision_model.encoder.layers.6.norm1.bias": "model-00004-of-00004.safetensors",
|
| 630 |
+
"vision_model.encoder.layers.6.norm1.weight": "model-00004-of-00004.safetensors",
|
| 631 |
+
"vision_model.encoder.layers.6.norm2.bias": "model-00004-of-00004.safetensors",
|
| 632 |
+
"vision_model.encoder.layers.6.norm2.weight": "model-00004-of-00004.safetensors",
|
| 633 |
+
"vision_model.encoder.layers.7.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 634 |
+
"vision_model.encoder.layers.7.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 635 |
+
"vision_model.encoder.layers.7.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 636 |
+
"vision_model.encoder.layers.7.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 637 |
+
"vision_model.encoder.layers.7.ls1": "model-00004-of-00004.safetensors",
|
| 638 |
+
"vision_model.encoder.layers.7.ls2": "model-00004-of-00004.safetensors",
|
| 639 |
+
"vision_model.encoder.layers.7.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 640 |
+
"vision_model.encoder.layers.7.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 641 |
+
"vision_model.encoder.layers.7.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 642 |
+
"vision_model.encoder.layers.7.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 643 |
+
"vision_model.encoder.layers.7.norm1.bias": "model-00004-of-00004.safetensors",
|
| 644 |
+
"vision_model.encoder.layers.7.norm1.weight": "model-00004-of-00004.safetensors",
|
| 645 |
+
"vision_model.encoder.layers.7.norm2.bias": "model-00004-of-00004.safetensors",
|
| 646 |
+
"vision_model.encoder.layers.7.norm2.weight": "model-00004-of-00004.safetensors",
|
| 647 |
+
"vision_model.encoder.layers.8.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 648 |
+
"vision_model.encoder.layers.8.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 649 |
+
"vision_model.encoder.layers.8.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 650 |
+
"vision_model.encoder.layers.8.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 651 |
+
"vision_model.encoder.layers.8.ls1": "model-00004-of-00004.safetensors",
|
| 652 |
+
"vision_model.encoder.layers.8.ls2": "model-00004-of-00004.safetensors",
|
| 653 |
+
"vision_model.encoder.layers.8.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 654 |
+
"vision_model.encoder.layers.8.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 655 |
+
"vision_model.encoder.layers.8.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 656 |
+
"vision_model.encoder.layers.8.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 657 |
+
"vision_model.encoder.layers.8.norm1.bias": "model-00004-of-00004.safetensors",
|
| 658 |
+
"vision_model.encoder.layers.8.norm1.weight": "model-00004-of-00004.safetensors",
|
| 659 |
+
"vision_model.encoder.layers.8.norm2.bias": "model-00004-of-00004.safetensors",
|
| 660 |
+
"vision_model.encoder.layers.8.norm2.weight": "model-00004-of-00004.safetensors",
|
| 661 |
+
"vision_model.encoder.layers.9.attn.proj.bias": "model-00004-of-00004.safetensors",
|
| 662 |
+
"vision_model.encoder.layers.9.attn.proj.weight": "model-00004-of-00004.safetensors",
|
| 663 |
+
"vision_model.encoder.layers.9.attn.qkv.bias": "model-00004-of-00004.safetensors",
|
| 664 |
+
"vision_model.encoder.layers.9.attn.qkv.weight": "model-00004-of-00004.safetensors",
|
| 665 |
+
"vision_model.encoder.layers.9.ls1": "model-00004-of-00004.safetensors",
|
| 666 |
+
"vision_model.encoder.layers.9.ls2": "model-00004-of-00004.safetensors",
|
| 667 |
+
"vision_model.encoder.layers.9.mlp.fc1.bias": "model-00004-of-00004.safetensors",
|
| 668 |
+
"vision_model.encoder.layers.9.mlp.fc1.weight": "model-00004-of-00004.safetensors",
|
| 669 |
+
"vision_model.encoder.layers.9.mlp.fc2.bias": "model-00004-of-00004.safetensors",
|
| 670 |
+
"vision_model.encoder.layers.9.mlp.fc2.weight": "model-00004-of-00004.safetensors",
|
| 671 |
+
"vision_model.encoder.layers.9.norm1.bias": "model-00004-of-00004.safetensors",
|
| 672 |
+
"vision_model.encoder.layers.9.norm1.weight": "model-00004-of-00004.safetensors",
|
| 673 |
+
"vision_model.encoder.layers.9.norm2.bias": "model-00004-of-00004.safetensors",
|
| 674 |
+
"vision_model.encoder.layers.9.norm2.weight": "model-00004-of-00004.safetensors"
|
| 675 |
+
}
|
| 676 |
+
}
|
modeling_intern_vit.py
ADDED
|
@@ -0,0 +1,428 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# --------------------------------------------------------
|
| 2 |
+
# InternVL
|
| 3 |
+
# Copyright (c) 2024 OpenGVLab
|
| 4 |
+
# Licensed under The MIT License [see LICENSE for details]
|
| 5 |
+
# --------------------------------------------------------
|
| 6 |
+
|
| 7 |
+
from typing import Optional, Tuple, Union
|
| 8 |
+
|
| 9 |
+
import torch
|
| 10 |
+
import torch.nn.functional as F
|
| 11 |
+
import torch.utils.checkpoint
|
| 12 |
+
from einops import rearrange
|
| 13 |
+
from timm.models.layers import DropPath
|
| 14 |
+
from torch import nn
|
| 15 |
+
from transformers.activations import ACT2FN
|
| 16 |
+
from transformers.modeling_outputs import (BaseModelOutput,
|
| 17 |
+
BaseModelOutputWithPooling)
|
| 18 |
+
from transformers.modeling_utils import PreTrainedModel
|
| 19 |
+
from transformers.utils import logging
|
| 20 |
+
|
| 21 |
+
from .configuration_intern_vit import InternVisionConfig
|
| 22 |
+
|
| 23 |
+
try:
|
| 24 |
+
from flash_attn.bert_padding import pad_input, unpad_input
|
| 25 |
+
from flash_attn.flash_attn_interface import \
|
| 26 |
+
flash_attn_varlen_qkvpacked_func
|
| 27 |
+
has_flash_attn = True
|
| 28 |
+
except:
|
| 29 |
+
print('FlashAttention2 is not installed.')
|
| 30 |
+
has_flash_attn = False
|
| 31 |
+
|
| 32 |
+
logger = logging.get_logger(__name__)
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
class FlashAttention(nn.Module):
|
| 36 |
+
"""Implement the scaled dot product attention with softmax.
|
| 37 |
+
Arguments
|
| 38 |
+
---------
|
| 39 |
+
softmax_scale: The temperature to use for the softmax attention.
|
| 40 |
+
(default: 1/sqrt(d_keys) where d_keys is computed at
|
| 41 |
+
runtime)
|
| 42 |
+
attention_dropout: The dropout rate to apply to the attention
|
| 43 |
+
(default: 0.0)
|
| 44 |
+
"""
|
| 45 |
+
|
| 46 |
+
def __init__(self, softmax_scale=None, attention_dropout=0.0, device=None, dtype=None):
|
| 47 |
+
super().__init__()
|
| 48 |
+
self.softmax_scale = softmax_scale
|
| 49 |
+
self.dropout_p = attention_dropout
|
| 50 |
+
|
| 51 |
+
def forward(self, qkv, key_padding_mask=None, causal=False, cu_seqlens=None,
|
| 52 |
+
max_s=None, need_weights=False):
|
| 53 |
+
"""Implements the multihead softmax attention.
|
| 54 |
+
Arguments
|
| 55 |
+
---------
|
| 56 |
+
qkv: The tensor containing the query, key, and value. (B, S, 3, H, D) if key_padding_mask is None
|
| 57 |
+
if unpadded: (nnz, 3, h, d)
|
| 58 |
+
key_padding_mask: a bool tensor of shape (B, S)
|
| 59 |
+
"""
|
| 60 |
+
assert not need_weights
|
| 61 |
+
assert qkv.dtype in [torch.float16, torch.bfloat16]
|
| 62 |
+
assert qkv.is_cuda
|
| 63 |
+
|
| 64 |
+
if cu_seqlens is None:
|
| 65 |
+
batch_size = qkv.shape[0]
|
| 66 |
+
seqlen = qkv.shape[1]
|
| 67 |
+
if key_padding_mask is None:
|
| 68 |
+
qkv = rearrange(qkv, 'b s ... -> (b s) ...')
|
| 69 |
+
max_s = seqlen
|
| 70 |
+
cu_seqlens = torch.arange(0, (batch_size + 1) * seqlen, step=seqlen, dtype=torch.int32,
|
| 71 |
+
device=qkv.device)
|
| 72 |
+
output = flash_attn_varlen_qkvpacked_func(
|
| 73 |
+
qkv, cu_seqlens, max_s, self.dropout_p if self.training else 0.0,
|
| 74 |
+
softmax_scale=self.softmax_scale, causal=causal
|
| 75 |
+
)
|
| 76 |
+
output = rearrange(output, '(b s) ... -> b s ...', b=batch_size)
|
| 77 |
+
else:
|
| 78 |
+
nheads = qkv.shape[-2]
|
| 79 |
+
x = rearrange(qkv, 'b s three h d -> b s (three h d)')
|
| 80 |
+
x_unpad, indices, cu_seqlens, max_s = unpad_input(x, key_padding_mask)
|
| 81 |
+
x_unpad = rearrange(x_unpad, 'nnz (three h d) -> nnz three h d', three=3, h=nheads)
|
| 82 |
+
output_unpad = flash_attn_varlen_qkvpacked_func(
|
| 83 |
+
x_unpad, cu_seqlens, max_s, self.dropout_p if self.training else 0.0,
|
| 84 |
+
softmax_scale=self.softmax_scale, causal=causal
|
| 85 |
+
)
|
| 86 |
+
output = rearrange(pad_input(rearrange(output_unpad, 'nnz h d -> nnz (h d)'),
|
| 87 |
+
indices, batch_size, seqlen),
|
| 88 |
+
'b s (h d) -> b s h d', h=nheads)
|
| 89 |
+
else:
|
| 90 |
+
assert max_s is not None
|
| 91 |
+
output = flash_attn_varlen_qkvpacked_func(
|
| 92 |
+
qkv, cu_seqlens, max_s, self.dropout_p if self.training else 0.0,
|
| 93 |
+
softmax_scale=self.softmax_scale, causal=causal
|
| 94 |
+
)
|
| 95 |
+
|
| 96 |
+
return output, None
|
| 97 |
+
|
| 98 |
+
|
| 99 |
+
|
| 100 |
+
class InternRMSNorm(nn.Module):
|
| 101 |
+
def __init__(self, hidden_size, eps=1e-6):
|
| 102 |
+
super().__init__()
|
| 103 |
+
self.weight = nn.Parameter(torch.ones(hidden_size))
|
| 104 |
+
self.variance_epsilon = eps
|
| 105 |
+
|
| 106 |
+
def forward(self, hidden_states):
|
| 107 |
+
input_dtype = hidden_states.dtype
|
| 108 |
+
hidden_states = hidden_states.to(torch.float32)
|
| 109 |
+
variance = hidden_states.pow(2).mean(-1, keepdim=True)
|
| 110 |
+
hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
|
| 111 |
+
return self.weight * hidden_states.to(input_dtype)
|
| 112 |
+
|
| 113 |
+
# try:
|
| 114 |
+
# from apex.normalization import FusedRMSNorm
|
| 115 |
+
|
| 116 |
+
# InternRMSNorm = FusedRMSNorm # noqa
|
| 117 |
+
|
| 118 |
+
# except ImportError:
|
| 119 |
+
# # using the normal InternRMSNorm
|
| 120 |
+
# pass
|
| 121 |
+
# except Exception:
|
| 122 |
+
# pass
|
| 123 |
+
|
| 124 |
+
|
| 125 |
+
NORM2FN = {
|
| 126 |
+
'rms_norm': InternRMSNorm,
|
| 127 |
+
'layer_norm': nn.LayerNorm,
|
| 128 |
+
}
|
| 129 |
+
|
| 130 |
+
|
| 131 |
+
class InternVisionEmbeddings(nn.Module):
|
| 132 |
+
def __init__(self, config: InternVisionConfig):
|
| 133 |
+
super().__init__()
|
| 134 |
+
self.config = config
|
| 135 |
+
self.embed_dim = config.hidden_size
|
| 136 |
+
self.image_size = config.image_size
|
| 137 |
+
self.patch_size = config.patch_size
|
| 138 |
+
|
| 139 |
+
self.class_embedding = nn.Parameter(
|
| 140 |
+
torch.randn(1, 1, self.embed_dim),
|
| 141 |
+
)
|
| 142 |
+
|
| 143 |
+
self.patch_embedding = nn.Conv2d(
|
| 144 |
+
in_channels=3, out_channels=self.embed_dim, kernel_size=self.patch_size, stride=self.patch_size
|
| 145 |
+
)
|
| 146 |
+
|
| 147 |
+
self.num_patches = (self.image_size // self.patch_size) ** 2
|
| 148 |
+
self.num_positions = self.num_patches + 1
|
| 149 |
+
|
| 150 |
+
self.position_embedding = nn.Parameter(torch.randn(1, self.num_positions, self.embed_dim))
|
| 151 |
+
|
| 152 |
+
def _get_pos_embed(self, pos_embed, H, W):
|
| 153 |
+
target_dtype = pos_embed.dtype
|
| 154 |
+
pos_embed = pos_embed.float().reshape(
|
| 155 |
+
1, self.image_size // self.patch_size, self.image_size // self.patch_size, -1).permute(0, 3, 1, 2)
|
| 156 |
+
pos_embed = F.interpolate(pos_embed, size=(H, W), mode='bicubic', align_corners=False). \
|
| 157 |
+
reshape(1, -1, H * W).permute(0, 2, 1).to(target_dtype)
|
| 158 |
+
return pos_embed
|
| 159 |
+
|
| 160 |
+
def forward(self, pixel_values: torch.FloatTensor) -> torch.Tensor:
|
| 161 |
+
target_dtype = self.patch_embedding.weight.dtype
|
| 162 |
+
patch_embeds = self.patch_embedding(pixel_values) # shape = [*, channel, width, height]
|
| 163 |
+
batch_size, _, height, width = patch_embeds.shape
|
| 164 |
+
patch_embeds = patch_embeds.flatten(2).transpose(1, 2)
|
| 165 |
+
class_embeds = self.class_embedding.expand(batch_size, 1, -1).to(target_dtype)
|
| 166 |
+
embeddings = torch.cat([class_embeds, patch_embeds], dim=1)
|
| 167 |
+
position_embedding = torch.cat([
|
| 168 |
+
self.position_embedding[:, :1, :],
|
| 169 |
+
self._get_pos_embed(self.position_embedding[:, 1:, :], height, width)
|
| 170 |
+
], dim=1)
|
| 171 |
+
embeddings = embeddings + position_embedding.to(target_dtype)
|
| 172 |
+
return embeddings
|
| 173 |
+
|
| 174 |
+
|
| 175 |
+
class InternAttention(nn.Module):
|
| 176 |
+
"""Multi-headed attention from 'Attention Is All You Need' paper"""
|
| 177 |
+
|
| 178 |
+
def __init__(self, config: InternVisionConfig):
|
| 179 |
+
super().__init__()
|
| 180 |
+
self.config = config
|
| 181 |
+
self.embed_dim = config.hidden_size
|
| 182 |
+
self.num_heads = config.num_attention_heads
|
| 183 |
+
self.use_flash_attn = config.use_flash_attn and has_flash_attn
|
| 184 |
+
if config.use_flash_attn and not has_flash_attn:
|
| 185 |
+
print('Warning: Flash Attention is not available, use_flash_attn is set to False.')
|
| 186 |
+
self.head_dim = self.embed_dim // self.num_heads
|
| 187 |
+
if self.head_dim * self.num_heads != self.embed_dim:
|
| 188 |
+
raise ValueError(
|
| 189 |
+
f'embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim} and `num_heads`:'
|
| 190 |
+
f' {self.num_heads}).'
|
| 191 |
+
)
|
| 192 |
+
|
| 193 |
+
self.scale = self.head_dim ** -0.5
|
| 194 |
+
self.qkv = nn.Linear(self.embed_dim, 3 * self.embed_dim, bias=config.qkv_bias)
|
| 195 |
+
self.attn_drop = nn.Dropout(config.attention_dropout)
|
| 196 |
+
self.proj_drop = nn.Dropout(config.dropout)
|
| 197 |
+
|
| 198 |
+
self.qk_normalization = config.qk_normalization
|
| 199 |
+
|
| 200 |
+
if self.qk_normalization:
|
| 201 |
+
self.q_norm = InternRMSNorm(self.embed_dim, eps=config.layer_norm_eps)
|
| 202 |
+
self.k_norm = InternRMSNorm(self.embed_dim, eps=config.layer_norm_eps)
|
| 203 |
+
|
| 204 |
+
if self.use_flash_attn:
|
| 205 |
+
self.inner_attn = FlashAttention(attention_dropout=config.attention_dropout)
|
| 206 |
+
self.proj = nn.Linear(self.embed_dim, self.embed_dim)
|
| 207 |
+
|
| 208 |
+
def _naive_attn(self, x):
|
| 209 |
+
B, N, C = x.shape
|
| 210 |
+
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
|
| 211 |
+
q, k, v = qkv.unbind(0) # make torchscript happy (cannot use tensor as tuple)
|
| 212 |
+
|
| 213 |
+
if self.qk_normalization:
|
| 214 |
+
B_, H_, N_, D_ = q.shape
|
| 215 |
+
q = self.q_norm(q.transpose(1, 2).flatten(-2, -1)).view(B_, N_, H_, D_).transpose(1, 2)
|
| 216 |
+
k = self.k_norm(k.transpose(1, 2).flatten(-2, -1)).view(B_, N_, H_, D_).transpose(1, 2)
|
| 217 |
+
|
| 218 |
+
attn = ((q * self.scale) @ k.transpose(-2, -1))
|
| 219 |
+
attn = attn.softmax(dim=-1)
|
| 220 |
+
attn = self.attn_drop(attn)
|
| 221 |
+
|
| 222 |
+
x = (attn @ v).transpose(1, 2).reshape(B, N, C)
|
| 223 |
+
x = self.proj(x)
|
| 224 |
+
x = self.proj_drop(x)
|
| 225 |
+
return x
|
| 226 |
+
|
| 227 |
+
def _flash_attn(self, x, key_padding_mask=None, need_weights=False):
|
| 228 |
+
qkv = self.qkv(x)
|
| 229 |
+
qkv = rearrange(qkv, 'b s (three h d) -> b s three h d', three=3, h=self.num_heads)
|
| 230 |
+
|
| 231 |
+
if self.qk_normalization:
|
| 232 |
+
q, k, v = qkv.unbind(2)
|
| 233 |
+
q = self.q_norm(q.flatten(-2, -1)).view(q.shape)
|
| 234 |
+
k = self.k_norm(k.flatten(-2, -1)).view(k.shape)
|
| 235 |
+
qkv = torch.stack([q, k, v], dim=2)
|
| 236 |
+
|
| 237 |
+
context, _ = self.inner_attn(
|
| 238 |
+
qkv, key_padding_mask=key_padding_mask, need_weights=need_weights, causal=False
|
| 239 |
+
)
|
| 240 |
+
outs = self.proj(rearrange(context, 'b s h d -> b s (h d)'))
|
| 241 |
+
outs = self.proj_drop(outs)
|
| 242 |
+
return outs
|
| 243 |
+
|
| 244 |
+
def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
|
| 245 |
+
x = self._naive_attn(hidden_states) if not self.use_flash_attn else self._flash_attn(hidden_states)
|
| 246 |
+
return x
|
| 247 |
+
|
| 248 |
+
|
| 249 |
+
class InternMLP(nn.Module):
|
| 250 |
+
def __init__(self, config: InternVisionConfig):
|
| 251 |
+
super().__init__()
|
| 252 |
+
self.config = config
|
| 253 |
+
self.act = ACT2FN[config.hidden_act]
|
| 254 |
+
self.fc1 = nn.Linear(config.hidden_size, config.intermediate_size)
|
| 255 |
+
self.fc2 = nn.Linear(config.intermediate_size, config.hidden_size)
|
| 256 |
+
|
| 257 |
+
def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
|
| 258 |
+
hidden_states = self.fc1(hidden_states)
|
| 259 |
+
hidden_states = self.act(hidden_states)
|
| 260 |
+
hidden_states = self.fc2(hidden_states)
|
| 261 |
+
return hidden_states
|
| 262 |
+
|
| 263 |
+
|
| 264 |
+
class InternVisionEncoderLayer(nn.Module):
|
| 265 |
+
def __init__(self, config: InternVisionConfig, drop_path_rate: float):
|
| 266 |
+
super().__init__()
|
| 267 |
+
self.embed_dim = config.hidden_size
|
| 268 |
+
self.intermediate_size = config.intermediate_size
|
| 269 |
+
self.norm_type = config.norm_type
|
| 270 |
+
|
| 271 |
+
self.attn = InternAttention(config)
|
| 272 |
+
self.mlp = InternMLP(config)
|
| 273 |
+
self.norm1 = NORM2FN[self.norm_type](self.embed_dim, eps=config.layer_norm_eps)
|
| 274 |
+
self.norm2 = NORM2FN[self.norm_type](self.embed_dim, eps=config.layer_norm_eps)
|
| 275 |
+
|
| 276 |
+
self.ls1 = nn.Parameter(config.initializer_factor * torch.ones(self.embed_dim))
|
| 277 |
+
self.ls2 = nn.Parameter(config.initializer_factor * torch.ones(self.embed_dim))
|
| 278 |
+
self.drop_path1 = DropPath(drop_path_rate) if drop_path_rate > 0. else nn.Identity()
|
| 279 |
+
self.drop_path2 = DropPath(drop_path_rate) if drop_path_rate > 0. else nn.Identity()
|
| 280 |
+
|
| 281 |
+
def forward(
|
| 282 |
+
self,
|
| 283 |
+
hidden_states: torch.Tensor,
|
| 284 |
+
) -> Tuple[torch.FloatTensor, Optional[torch.FloatTensor], Optional[Tuple[torch.FloatTensor]]]:
|
| 285 |
+
"""
|
| 286 |
+
Args:
|
| 287 |
+
hidden_states (`Tuple[torch.FloatTensor, Optional[torch.FloatTensor]]`): input to the layer of shape `(batch, seq_len, embed_dim)`
|
| 288 |
+
"""
|
| 289 |
+
hidden_states = hidden_states + self.drop_path1(self.attn(self.norm1(hidden_states).to(hidden_states.dtype)) * self.ls1)
|
| 290 |
+
|
| 291 |
+
hidden_states = hidden_states + self.drop_path2(self.mlp(self.norm2(hidden_states).to(hidden_states.dtype)) * self.ls2)
|
| 292 |
+
|
| 293 |
+
return hidden_states
|
| 294 |
+
|
| 295 |
+
|
| 296 |
+
class InternVisionEncoder(nn.Module):
|
| 297 |
+
"""
|
| 298 |
+
Transformer encoder consisting of `config.num_hidden_layers` self attention layers. Each layer is a
|
| 299 |
+
[`InternEncoderLayer`].
|
| 300 |
+
|
| 301 |
+
Args:
|
| 302 |
+
config (`InternConfig`):
|
| 303 |
+
The corresponding vision configuration for the `InternEncoder`.
|
| 304 |
+
"""
|
| 305 |
+
|
| 306 |
+
def __init__(self, config: InternVisionConfig):
|
| 307 |
+
super().__init__()
|
| 308 |
+
self.config = config
|
| 309 |
+
# stochastic depth decay rule
|
| 310 |
+
dpr = [x.item() for x in torch.linspace(0, config.drop_path_rate, config.num_hidden_layers)]
|
| 311 |
+
self.layers = nn.ModuleList([
|
| 312 |
+
InternVisionEncoderLayer(config, dpr[idx]) for idx in range(config.num_hidden_layers)])
|
| 313 |
+
self.gradient_checkpointing = True
|
| 314 |
+
|
| 315 |
+
def forward(
|
| 316 |
+
self,
|
| 317 |
+
inputs_embeds,
|
| 318 |
+
output_hidden_states: Optional[bool] = None,
|
| 319 |
+
return_dict: Optional[bool] = None,
|
| 320 |
+
) -> Union[Tuple, BaseModelOutput]:
|
| 321 |
+
r"""
|
| 322 |
+
Args:
|
| 323 |
+
inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
|
| 324 |
+
Embedded representation of the inputs. Should be float, not int tokens.
|
| 325 |
+
output_hidden_states (`bool`, *optional*):
|
| 326 |
+
Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors
|
| 327 |
+
for more detail.
|
| 328 |
+
return_dict (`bool`, *optional*):
|
| 329 |
+
Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
|
| 330 |
+
"""
|
| 331 |
+
output_hidden_states = (
|
| 332 |
+
output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
|
| 333 |
+
)
|
| 334 |
+
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
| 335 |
+
|
| 336 |
+
encoder_states = () if output_hidden_states else None
|
| 337 |
+
hidden_states = inputs_embeds
|
| 338 |
+
|
| 339 |
+
for idx, encoder_layer in enumerate(self.layers):
|
| 340 |
+
if output_hidden_states:
|
| 341 |
+
encoder_states = encoder_states + (hidden_states,)
|
| 342 |
+
if self.gradient_checkpointing and self.training:
|
| 343 |
+
layer_outputs = torch.utils.checkpoint.checkpoint(
|
| 344 |
+
encoder_layer,
|
| 345 |
+
hidden_states)
|
| 346 |
+
else:
|
| 347 |
+
layer_outputs = encoder_layer(
|
| 348 |
+
hidden_states,
|
| 349 |
+
)
|
| 350 |
+
hidden_states = layer_outputs
|
| 351 |
+
|
| 352 |
+
if output_hidden_states:
|
| 353 |
+
encoder_states = encoder_states + (hidden_states,)
|
| 354 |
+
|
| 355 |
+
if not return_dict:
|
| 356 |
+
return tuple(v for v in [hidden_states, encoder_states] if v is not None)
|
| 357 |
+
return BaseModelOutput(
|
| 358 |
+
last_hidden_state=hidden_states, hidden_states=encoder_states
|
| 359 |
+
)
|
| 360 |
+
|
| 361 |
+
|
| 362 |
+
class InternVisionModel(PreTrainedModel):
|
| 363 |
+
main_input_name = 'pixel_values'
|
| 364 |
+
_supports_flash_attn_2 = True
|
| 365 |
+
config_class = InternVisionConfig
|
| 366 |
+
_no_split_modules = ['InternVisionEncoderLayer']
|
| 367 |
+
|
| 368 |
+
def __init__(self, config: InternVisionConfig):
|
| 369 |
+
super().__init__(config)
|
| 370 |
+
self.config = config
|
| 371 |
+
|
| 372 |
+
self.embeddings = InternVisionEmbeddings(config)
|
| 373 |
+
self.encoder = InternVisionEncoder(config)
|
| 374 |
+
|
| 375 |
+
def resize_pos_embeddings(self, old_size, new_size, patch_size):
|
| 376 |
+
pos_emb = self.embeddings.position_embedding
|
| 377 |
+
_, num_positions, embed_dim = pos_emb.shape
|
| 378 |
+
cls_emb = pos_emb[:, :1, :]
|
| 379 |
+
pos_emb = pos_emb[:, 1:, :].reshape(1, old_size // patch_size, old_size // patch_size, -1).permute(0, 3, 1, 2)
|
| 380 |
+
pos_emb = F.interpolate(pos_emb.float(), size=new_size // patch_size, mode='bicubic', align_corners=False)
|
| 381 |
+
pos_emb = pos_emb.to(cls_emb.dtype).reshape(1, embed_dim, -1).permute(0, 2, 1)
|
| 382 |
+
pos_emb = torch.cat([cls_emb, pos_emb], dim=1)
|
| 383 |
+
self.embeddings.position_embedding = nn.Parameter(pos_emb)
|
| 384 |
+
self.embeddings.image_size = new_size
|
| 385 |
+
logger.info('Resized position embeddings from {} to {}'.format(old_size, new_size))
|
| 386 |
+
|
| 387 |
+
def get_input_embeddings(self):
|
| 388 |
+
return self.embeddings
|
| 389 |
+
|
| 390 |
+
def forward(
|
| 391 |
+
self,
|
| 392 |
+
pixel_values: Optional[torch.FloatTensor] = None,
|
| 393 |
+
output_hidden_states: Optional[bool] = None,
|
| 394 |
+
return_dict: Optional[bool] = None,
|
| 395 |
+
pixel_embeds: Optional[torch.FloatTensor] = None,
|
| 396 |
+
) -> Union[Tuple, BaseModelOutputWithPooling]:
|
| 397 |
+
output_hidden_states = (
|
| 398 |
+
output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
|
| 399 |
+
)
|
| 400 |
+
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
| 401 |
+
|
| 402 |
+
if pixel_values is None and pixel_embeds is None:
|
| 403 |
+
raise ValueError('You have to specify pixel_values or pixel_embeds')
|
| 404 |
+
|
| 405 |
+
if pixel_embeds is not None:
|
| 406 |
+
hidden_states = pixel_embeds
|
| 407 |
+
else:
|
| 408 |
+
if len(pixel_values.shape) == 4:
|
| 409 |
+
hidden_states = self.embeddings(pixel_values)
|
| 410 |
+
else:
|
| 411 |
+
raise ValueError(f'wrong pixel_values size: {pixel_values.shape}')
|
| 412 |
+
encoder_outputs = self.encoder(
|
| 413 |
+
inputs_embeds=hidden_states,
|
| 414 |
+
output_hidden_states=output_hidden_states,
|
| 415 |
+
return_dict=return_dict,
|
| 416 |
+
)
|
| 417 |
+
last_hidden_state = encoder_outputs.last_hidden_state
|
| 418 |
+
pooled_output = last_hidden_state[:, 0, :]
|
| 419 |
+
|
| 420 |
+
if not return_dict:
|
| 421 |
+
return (last_hidden_state, pooled_output) + encoder_outputs[1:]
|
| 422 |
+
|
| 423 |
+
return BaseModelOutputWithPooling(
|
| 424 |
+
last_hidden_state=last_hidden_state,
|
| 425 |
+
pooler_output=pooled_output,
|
| 426 |
+
hidden_states=encoder_outputs.hidden_states,
|
| 427 |
+
attentions=encoder_outputs.attentions,
|
| 428 |
+
)
|
modeling_qianfanvl_chat.py
ADDED
|
@@ -0,0 +1,394 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Copyright (c) 2025 Qianfan
|
| 2 |
+
# Licensed under the MIT License. See LICENSE file in the project root for full license information.
|
| 3 |
+
import warnings
|
| 4 |
+
from typing import List, Optional, Tuple, Union
|
| 5 |
+
from transformers import AutoTokenizer
|
| 6 |
+
import torch.distributed as dist
|
| 7 |
+
import torch.utils.checkpoint
|
| 8 |
+
import transformers
|
| 9 |
+
from .conversation import get_conv_template
|
| 10 |
+
from torch import nn
|
| 11 |
+
from torch.nn import CrossEntropyLoss
|
| 12 |
+
from transformers import (AutoModel, GenerationConfig, LlamaForCausalLM,
|
| 13 |
+
LlamaTokenizer)
|
| 14 |
+
from transformers.modeling_outputs import CausalLMOutputWithPast
|
| 15 |
+
from transformers.modeling_utils import PreTrainedModel
|
| 16 |
+
from transformers.utils import ModelOutput, logging
|
| 17 |
+
|
| 18 |
+
from .configuration_qianfanvl_chat import QianfanVLChatConfig
|
| 19 |
+
from .modeling_intern_vit import InternVisionModel, has_flash_attn
|
| 20 |
+
import difflib
|
| 21 |
+
|
| 22 |
+
logger = logging.get_logger(__name__)
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
def version_cmp(v1, v2, op='eq'):
|
| 26 |
+
import operator
|
| 27 |
+
|
| 28 |
+
from packaging import version
|
| 29 |
+
op_func = getattr(operator, op)
|
| 30 |
+
return op_func(version.parse(v1), version.parse(v2))
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
class QianfanVLChatModel(PreTrainedModel):
|
| 34 |
+
config_class = QianfanVLChatConfig
|
| 35 |
+
main_input_name = 'pixel_values'
|
| 36 |
+
base_model_prefix = 'language_model'
|
| 37 |
+
_no_split_modules = ['InternVisionModel', 'LlamaDecoderLayer']
|
| 38 |
+
_supports_flash_attn_2 = True
|
| 39 |
+
supports_gradient_checkpointing = True
|
| 40 |
+
|
| 41 |
+
def __init__(self, config: QianfanVLChatConfig, vision_model=None, language_model=None, use_flash_attn=True):
|
| 42 |
+
super().__init__(config)
|
| 43 |
+
|
| 44 |
+
assert version_cmp(transformers.__version__, '4.37.0', 'ge')
|
| 45 |
+
image_size = config.force_image_size or config.vision_config.image_size
|
| 46 |
+
patch_size = config.vision_config.patch_size
|
| 47 |
+
self.patch_size = patch_size
|
| 48 |
+
self.select_layer = config.select_layer
|
| 49 |
+
self.template = config.template
|
| 50 |
+
self.num_image_token = int((image_size // patch_size) ** 2 * (config.downsample_ratio ** 2))
|
| 51 |
+
self.downsample_ratio = config.downsample_ratio
|
| 52 |
+
self.ps_version = config.ps_version
|
| 53 |
+
self.llm_arch_name = config.llm_config.architectures[0]
|
| 54 |
+
# Enable Flash Attention if supported, otherwise fall back to eager attention.
|
| 55 |
+
use_flash_attn = False # use_flash_attn if has_flash_attn else False
|
| 56 |
+
config.vision_config.use_flash_attn = True if use_flash_attn else False
|
| 57 |
+
config.llm_config.attn_implementation = 'flash_attention_2' if use_flash_attn else 'eager'
|
| 58 |
+
|
| 59 |
+
if vision_model is not None:
|
| 60 |
+
self.vision_model = vision_model
|
| 61 |
+
else:
|
| 62 |
+
self.vision_model = InternVisionModel(config.vision_config)
|
| 63 |
+
if language_model is not None:
|
| 64 |
+
self.language_model = language_model
|
| 65 |
+
else:
|
| 66 |
+
self.language_model = LlamaForCausalLM(config.llm_config)
|
| 67 |
+
|
| 68 |
+
vit_hidden_size = config.vision_config.hidden_size
|
| 69 |
+
llm_hidden_size = config.llm_config.hidden_size
|
| 70 |
+
|
| 71 |
+
self.mlp1 = nn.Sequential(
|
| 72 |
+
nn.LayerNorm(vit_hidden_size * int(1 / self.downsample_ratio) ** 2),
|
| 73 |
+
nn.Linear(vit_hidden_size * int(1 / self.downsample_ratio) ** 2, llm_hidden_size),
|
| 74 |
+
nn.GELU(),
|
| 75 |
+
nn.Linear(llm_hidden_size, llm_hidden_size)
|
| 76 |
+
)
|
| 77 |
+
|
| 78 |
+
self.img_context_token_id = None
|
| 79 |
+
self.conv_template = get_conv_template(self.template)
|
| 80 |
+
if hasattr(config, 'system_message'):
|
| 81 |
+
self.system_message = config.system_message
|
| 82 |
+
else:
|
| 83 |
+
self.system_message = self.conv_template.system_message
|
| 84 |
+
self.num_samples = 0
|
| 85 |
+
|
| 86 |
+
def forward(
|
| 87 |
+
self,
|
| 88 |
+
pixel_values: torch.FloatTensor,
|
| 89 |
+
input_ids: torch.LongTensor = None,
|
| 90 |
+
attention_mask: Optional[torch.Tensor] = None,
|
| 91 |
+
position_ids: Optional[torch.LongTensor] = None,
|
| 92 |
+
image_flags: Optional[torch.LongTensor] = None,
|
| 93 |
+
past_key_values: Optional[List[torch.FloatTensor]] = None,
|
| 94 |
+
labels: Optional[torch.LongTensor] = None,
|
| 95 |
+
use_cache: Optional[bool] = None,
|
| 96 |
+
output_attentions: Optional[bool] = None,
|
| 97 |
+
output_hidden_states: Optional[bool] = None,
|
| 98 |
+
return_dict: Optional[bool] = None,
|
| 99 |
+
statistics: Optional[torch.LongTensor] = None,
|
| 100 |
+
loss_weight: Optional[List] = None,
|
| 101 |
+
loss_reduction_all_gather: Optional[bool] = False,
|
| 102 |
+
) -> Union[Tuple, CausalLMOutputWithPast]:
|
| 103 |
+
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
|
| 104 |
+
|
| 105 |
+
image_flags = image_flags.squeeze(-1)
|
| 106 |
+
input_embeds = self.language_model.get_input_embeddings()(input_ids).clone()
|
| 107 |
+
vit_embeds = self.extract_feature(pixel_values)
|
| 108 |
+
vit_embeds = vit_embeds[image_flags == 1]
|
| 109 |
+
vit_batch_size = pixel_values.shape[0]
|
| 110 |
+
B, N, C = input_embeds.shape
|
| 111 |
+
input_embeds = input_embeds.reshape(B * N, C)
|
| 112 |
+
|
| 113 |
+
if torch.distributed.is_initialized() and torch.distributed.get_rank() == 0:
|
| 114 |
+
print(f'dynamic ViT batch size: {vit_batch_size}, images per sample: {vit_batch_size / B}, dynamic token length: {N}')
|
| 115 |
+
if statistics is not None:
|
| 116 |
+
num_samples, num_padding_tokens, num_padding_images = statistics.tolist()
|
| 117 |
+
self.num_samples += num_samples
|
| 118 |
+
print(f'total_samples={self.num_samples}, {num_samples=}, {num_padding_tokens=}, {num_padding_images=}')
|
| 119 |
+
|
| 120 |
+
input_ids = input_ids.reshape(B * N)
|
| 121 |
+
selected = (input_ids == self.img_context_token_id)
|
| 122 |
+
try:
|
| 123 |
+
input_embeds[selected] = input_embeds[selected] * 0.0 + vit_embeds.reshape(-1, C)
|
| 124 |
+
ignore_flag = False
|
| 125 |
+
except Exception as e:
|
| 126 |
+
vit_embeds = vit_embeds.reshape(-1, C)
|
| 127 |
+
print(f'warning: {e}, input_embeds[selected].shape={input_embeds[selected].shape}, '
|
| 128 |
+
f'vit_embeds.shape={vit_embeds.shape}')
|
| 129 |
+
n_token = selected.sum()
|
| 130 |
+
input_embeds[selected] = input_embeds[selected] * 0.0 + vit_embeds[:n_token]
|
| 131 |
+
ignore_flag = True
|
| 132 |
+
|
| 133 |
+
input_embeds = input_embeds.reshape(B, N, C)
|
| 134 |
+
|
| 135 |
+
outputs = self.language_model(
|
| 136 |
+
inputs_embeds=input_embeds,
|
| 137 |
+
attention_mask=attention_mask,
|
| 138 |
+
position_ids=position_ids,
|
| 139 |
+
past_key_values=past_key_values,
|
| 140 |
+
use_cache=use_cache,
|
| 141 |
+
output_attentions=output_attentions,
|
| 142 |
+
output_hidden_states=output_hidden_states,
|
| 143 |
+
return_dict=return_dict,
|
| 144 |
+
)
|
| 145 |
+
logits = outputs.logits
|
| 146 |
+
|
| 147 |
+
loss = None
|
| 148 |
+
if labels is not None and loss_weight is not None:
|
| 149 |
+
loss_weight = torch.tensor(loss_weight, dtype=torch.float32, device=labels.device)
|
| 150 |
+
# Shift so that tokens < n predict n
|
| 151 |
+
shift_logits = logits[..., :-1, :].contiguous()
|
| 152 |
+
shift_labels = labels[..., 1:].contiguous()
|
| 153 |
+
shift_weights = loss_weight[..., 1:].contiguous()
|
| 154 |
+
# Flatten the tokens
|
| 155 |
+
loss_fct = CrossEntropyLoss(reduction='none')
|
| 156 |
+
shift_logits = shift_logits.view(-1, self.language_model.config.vocab_size)
|
| 157 |
+
shift_labels = shift_labels.view(-1)
|
| 158 |
+
shift_weights = shift_weights.view(-1)
|
| 159 |
+
# Enable model parallelism
|
| 160 |
+
shift_labels = shift_labels.to(shift_logits.device)
|
| 161 |
+
shift_weights = shift_weights.to(shift_logits.device)
|
| 162 |
+
loss = loss_fct(shift_logits, shift_labels)
|
| 163 |
+
|
| 164 |
+
shift_weights_sum = shift_weights.sum()
|
| 165 |
+
if loss_reduction_all_gather:
|
| 166 |
+
dist.all_reduce(shift_weights_sum, op=dist.ReduceOp.AVG)
|
| 167 |
+
|
| 168 |
+
loss = loss * shift_weights
|
| 169 |
+
loss = loss.sum() / shift_weights_sum
|
| 170 |
+
if ignore_flag:
|
| 171 |
+
loss = loss * 0.0
|
| 172 |
+
elif labels is not None:
|
| 173 |
+
# Shift so that tokens < n predict n
|
| 174 |
+
shift_logits = logits[..., :-1, :].contiguous()
|
| 175 |
+
shift_labels = labels[..., 1:].contiguous()
|
| 176 |
+
# Flatten the tokens
|
| 177 |
+
loss_fct = CrossEntropyLoss()
|
| 178 |
+
shift_logits = shift_logits.view(-1, self.language_model.config.vocab_size)
|
| 179 |
+
shift_labels = shift_labels.view(-1)
|
| 180 |
+
# Enable model parallelism
|
| 181 |
+
shift_labels = shift_labels.to(shift_logits.device)
|
| 182 |
+
loss = loss_fct(shift_logits, shift_labels)
|
| 183 |
+
if ignore_flag:
|
| 184 |
+
loss = loss * 0.0
|
| 185 |
+
|
| 186 |
+
if not return_dict:
|
| 187 |
+
output = (logits,) + outputs[1:]
|
| 188 |
+
return (loss,) + output if loss is not None else output
|
| 189 |
+
|
| 190 |
+
return CausalLMOutputWithPast(
|
| 191 |
+
loss=loss,
|
| 192 |
+
logits=logits,
|
| 193 |
+
past_key_values=outputs.past_key_values,
|
| 194 |
+
hidden_states=outputs.hidden_states,
|
| 195 |
+
attentions=outputs.attentions,
|
| 196 |
+
)
|
| 197 |
+
|
| 198 |
+
def pixel_shuffle(self, x, scale_factor=0.5):
|
| 199 |
+
n, w, h, c = x.size()
|
| 200 |
+
# N, W, H, C --> N, W, H * scale, C // scale
|
| 201 |
+
x = x.view(n, w, int(h * scale_factor), int(c / scale_factor))
|
| 202 |
+
# N, W, H * scale, C // scale --> N, H * scale, W, C // scale
|
| 203 |
+
x = x.permute(0, 2, 1, 3).contiguous()
|
| 204 |
+
# N, H * scale, W, C // scale --> N, H * scale, W * scale, C // (scale ** 2)
|
| 205 |
+
x = x.view(n, int(h * scale_factor), int(w * scale_factor),
|
| 206 |
+
int(c / (scale_factor * scale_factor)))
|
| 207 |
+
if self.ps_version == 'v1':
|
| 208 |
+
warnings.warn("In ps_version 'v1', the height and width have not been swapped back, "
|
| 209 |
+
'which results in a transposed image.')
|
| 210 |
+
else:
|
| 211 |
+
x = x.permute(0, 2, 1, 3).contiguous()
|
| 212 |
+
return x
|
| 213 |
+
|
| 214 |
+
def extract_feature(self, pixel_values):
|
| 215 |
+
if self.select_layer == -1:
|
| 216 |
+
vit_embeds = self.vision_model(
|
| 217 |
+
pixel_values=pixel_values,
|
| 218 |
+
output_hidden_states=False,
|
| 219 |
+
return_dict=True).last_hidden_state
|
| 220 |
+
else:
|
| 221 |
+
vit_embeds = self.vision_model(
|
| 222 |
+
pixel_values=pixel_values,
|
| 223 |
+
output_hidden_states=True,
|
| 224 |
+
return_dict=True).hidden_states[self.select_layer]
|
| 225 |
+
vit_embeds = vit_embeds[:, 1:, :]
|
| 226 |
+
h = w = int(vit_embeds.shape[1] ** 0.5)
|
| 227 |
+
vit_embeds = vit_embeds.reshape(vit_embeds.shape[0], h, w, -1)
|
| 228 |
+
vit_embeds = self.pixel_shuffle(vit_embeds, scale_factor=self.downsample_ratio)
|
| 229 |
+
vit_embeds = vit_embeds.reshape(vit_embeds.shape[0], -1, vit_embeds.shape[-1])
|
| 230 |
+
vit_embeds = self.mlp1(vit_embeds)
|
| 231 |
+
return vit_embeds
|
| 232 |
+
|
| 233 |
+
def batch_chat(self, tokenizer, pixel_values, questions, generation_config, num_patches_list=None,
|
| 234 |
+
history=None, return_history=False, IMG_START_TOKEN='<img>', IMG_END_TOKEN='</img>',
|
| 235 |
+
IMG_CONTEXT_TOKEN='<IMG_CONTEXT>', verbose=False, image_counts=None):
|
| 236 |
+
if history is not None or return_history:
|
| 237 |
+
print('Now multi-turn chat is not supported in batch_chat.')
|
| 238 |
+
raise NotImplementedError
|
| 239 |
+
|
| 240 |
+
if image_counts is not None:
|
| 241 |
+
num_patches_list = image_counts
|
| 242 |
+
print('Warning: `image_counts` is deprecated. Please use `num_patches_list` instead.')
|
| 243 |
+
|
| 244 |
+
img_context_token_id = tokenizer.convert_tokens_to_ids(IMG_CONTEXT_TOKEN)
|
| 245 |
+
self.img_context_token_id = img_context_token_id
|
| 246 |
+
|
| 247 |
+
if verbose and pixel_values is not None:
|
| 248 |
+
image_bs = pixel_values.shape[0]
|
| 249 |
+
print(f'dynamic ViT batch size: {image_bs}')
|
| 250 |
+
|
| 251 |
+
queries = []
|
| 252 |
+
for idx, num_patches in enumerate(num_patches_list):
|
| 253 |
+
question = questions[idx]
|
| 254 |
+
if pixel_values is not None and '<image>' not in question:
|
| 255 |
+
question = '<image>' + question
|
| 256 |
+
template = get_conv_template(self.template)
|
| 257 |
+
template.system_message = self.system_message
|
| 258 |
+
template.append_message(template.roles[0], question)
|
| 259 |
+
template.append_message(template.roles[1], None)
|
| 260 |
+
query = template.get_prompt()
|
| 261 |
+
|
| 262 |
+
image_tokens = IMG_START_TOKEN + IMG_CONTEXT_TOKEN * self.num_image_token * num_patches + IMG_END_TOKEN
|
| 263 |
+
query = query.replace('<image>', image_tokens, 1)
|
| 264 |
+
queries.append(query)
|
| 265 |
+
|
| 266 |
+
tokenizer.padding_side = 'left'
|
| 267 |
+
model_inputs = tokenizer(queries, return_tensors='pt', padding=True)
|
| 268 |
+
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
| 269 |
+
input_ids = model_inputs['input_ids'].to(device)
|
| 270 |
+
attention_mask = model_inputs['attention_mask'].to(device)
|
| 271 |
+
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep.strip())
|
| 272 |
+
generation_config['eos_token_id'] = eos_token_id
|
| 273 |
+
#generation_config['eos_token_id'] = 181896
|
| 274 |
+
generation_output = self.generate(
|
| 275 |
+
pixel_values=pixel_values,
|
| 276 |
+
input_ids=input_ids,
|
| 277 |
+
attention_mask=attention_mask,
|
| 278 |
+
**generation_config
|
| 279 |
+
)
|
| 280 |
+
responses = tokenizer.batch_decode(generation_output, skip_special_tokens=False)
|
| 281 |
+
responses = [response.split(template.sep.strip())[0].strip() for response in responses]
|
| 282 |
+
return responses
|
| 283 |
+
|
| 284 |
+
def chat(self, tokenizer, pixel_values, question, generation_config, history=None, return_history=False,
|
| 285 |
+
num_patches_list=None, IMG_START_TOKEN='<img>', IMG_END_TOKEN='</img>', IMG_CONTEXT_TOKEN='<IMG_CONTEXT>',
|
| 286 |
+
verbose=False, debug=False):
|
| 287 |
+
|
| 288 |
+
if history is None and pixel_values is not None and '<image>' not in question:
|
| 289 |
+
question = '<image>' + question
|
| 290 |
+
|
| 291 |
+
if num_patches_list is None:
|
| 292 |
+
num_patches_list = [pixel_values.shape[0]] if pixel_values is not None else []
|
| 293 |
+
assert pixel_values is None or len(pixel_values) == sum(num_patches_list)
|
| 294 |
+
|
| 295 |
+
img_context_token_id = tokenizer.convert_tokens_to_ids(IMG_CONTEXT_TOKEN)
|
| 296 |
+
self.img_context_token_id = img_context_token_id
|
| 297 |
+
|
| 298 |
+
template = get_conv_template(self.template)
|
| 299 |
+
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep.strip())
|
| 300 |
+
|
| 301 |
+
history = [] if history is None else history
|
| 302 |
+
for (old_question, old_answer) in history:
|
| 303 |
+
template.append_message(template.roles[0], old_question)
|
| 304 |
+
template.append_message(template.roles[1], old_answer)
|
| 305 |
+
template.append_message(template.roles[0], question)
|
| 306 |
+
template.append_message(template.roles[1], None)
|
| 307 |
+
query = template.get_prompt()
|
| 308 |
+
|
| 309 |
+
if verbose and pixel_values is not None:
|
| 310 |
+
image_bs = pixel_values.shape[0]
|
| 311 |
+
print(f'dynamic ViT batch size: {image_bs}')
|
| 312 |
+
|
| 313 |
+
for num_patches in num_patches_list:
|
| 314 |
+
image_tokens = IMG_START_TOKEN + IMG_CONTEXT_TOKEN * self.num_image_token * num_patches + IMG_END_TOKEN
|
| 315 |
+
query = query.replace('<image>', image_tokens, 1)
|
| 316 |
+
|
| 317 |
+
model_inputs = tokenizer(query, return_tensors='pt')
|
| 318 |
+
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
| 319 |
+
input_ids = model_inputs['input_ids'].to(device)
|
| 320 |
+
attention_mask = model_inputs['attention_mask'].to(device)
|
| 321 |
+
generation_config['eos_token_id'] = eos_token_id
|
| 322 |
+
|
| 323 |
+
generation_output = self.generate(
|
| 324 |
+
pixel_values=pixel_values,
|
| 325 |
+
input_ids=input_ids,
|
| 326 |
+
attention_mask=attention_mask,
|
| 327 |
+
**generation_config
|
| 328 |
+
)
|
| 329 |
+
|
| 330 |
+
if debug:
|
| 331 |
+
return input_ids, generation_output
|
| 332 |
+
response = tokenizer.batch_decode(generation_output, skip_special_tokens=False)[0]
|
| 333 |
+
response = response.split(template.sep.strip())[0].strip()
|
| 334 |
+
history.append((question, response))
|
| 335 |
+
if return_history:
|
| 336 |
+
return response, history
|
| 337 |
+
else:
|
| 338 |
+
query_to_print = query.replace(IMG_CONTEXT_TOKEN, '')
|
| 339 |
+
query_to_print = query_to_print.replace(f'{IMG_START_TOKEN}{IMG_END_TOKEN}', '<image>')
|
| 340 |
+
if verbose:
|
| 341 |
+
print(query_to_print, response)
|
| 342 |
+
return response
|
| 343 |
+
|
| 344 |
+
@torch.no_grad()
|
| 345 |
+
def generate(
|
| 346 |
+
self,
|
| 347 |
+
pixel_values: Optional[torch.FloatTensor] = None,
|
| 348 |
+
input_ids: Optional[torch.FloatTensor] = None,
|
| 349 |
+
attention_mask: Optional[torch.LongTensor] = None,
|
| 350 |
+
visual_features: Optional[torch.FloatTensor] = None,
|
| 351 |
+
generation_config: Optional[GenerationConfig] = None,
|
| 352 |
+
output_hidden_states: Optional[bool] = None,
|
| 353 |
+
**generate_kwargs,
|
| 354 |
+
):
|
| 355 |
+
|
| 356 |
+
assert self.img_context_token_id is not None
|
| 357 |
+
if pixel_values is not None:
|
| 358 |
+
if visual_features is not None:
|
| 359 |
+
vit_embeds = visual_features
|
| 360 |
+
else:
|
| 361 |
+
vit_embeds = self.extract_feature(pixel_values)
|
| 362 |
+
#print("ViTEmbedding shape:", vit_embeds.shape)
|
| 363 |
+
input_embeds = self.language_model.get_input_embeddings()(input_ids)
|
| 364 |
+
B, N, C = input_embeds.shape
|
| 365 |
+
input_embeds = input_embeds.reshape(B * N, C)
|
| 366 |
+
input_ids = input_ids.reshape(B * N)
|
| 367 |
+
selected = (input_ids == self.img_context_token_id)
|
| 368 |
+
assert selected.sum() != 0
|
| 369 |
+
input_embeds[selected] = vit_embeds.reshape(-1, C).to(input_embeds.device)
|
| 370 |
+
|
| 371 |
+
input_embeds = input_embeds.reshape(B, N, C)
|
| 372 |
+
else:
|
| 373 |
+
input_embeds = self.language_model.get_input_embeddings()(input_ids)
|
| 374 |
+
|
| 375 |
+
outputs = self.language_model.generate(
|
| 376 |
+
inputs_embeds=input_embeds,
|
| 377 |
+
attention_mask=attention_mask,
|
| 378 |
+
generation_config=generation_config,
|
| 379 |
+
output_hidden_states=output_hidden_states,
|
| 380 |
+
use_cache=True,
|
| 381 |
+
**generate_kwargs,
|
| 382 |
+
)
|
| 383 |
+
|
| 384 |
+
return outputs
|
| 385 |
+
|
| 386 |
+
@property
|
| 387 |
+
def lm_head(self):
|
| 388 |
+
return self.language_model.get_output_embeddings()
|
| 389 |
+
|
| 390 |
+
def get_input_embeddings(self):
|
| 391 |
+
return self.language_model.get_input_embeddings()
|
| 392 |
+
|
| 393 |
+
def get_output_embeddings(self):
|
| 394 |
+
return self.language_model.get_output_embeddings()
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|begin_of_text|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|end_of_text|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|end_of_text|>"
|
| 17 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc9a1340b0cfcda70bee91abbf2adff4dc0dd4c9b2f5f9b9e86926fa0fe92fc0
|
| 3 |
+
size 12264545
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,1120 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_eos_token": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"181887": {
|
| 5 |
+
"content": "<|begin_of_text|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"181888": {
|
| 13 |
+
"content": "<|end_of_text|>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"181889": {
|
| 21 |
+
"content": "<|reserved_special_token_0|>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"181890": {
|
| 29 |
+
"content": "<|reserved_special_token_1|>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"181891": {
|
| 37 |
+
"content": "<|reserved_special_token_2|>",
|
| 38 |
+
"lstrip": false,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
},
|
| 44 |
+
"181892": {
|
| 45 |
+
"content": "<|reserved_special_token_3|>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false,
|
| 50 |
+
"special": true
|
| 51 |
+
},
|
| 52 |
+
"181893": {
|
| 53 |
+
"content": "<|start_header_id|>",
|
| 54 |
+
"lstrip": false,
|
| 55 |
+
"normalized": false,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false,
|
| 58 |
+
"special": true
|
| 59 |
+
},
|
| 60 |
+
"181894": {
|
| 61 |
+
"content": "<|end_header_id|>",
|
| 62 |
+
"lstrip": false,
|
| 63 |
+
"normalized": false,
|
| 64 |
+
"rstrip": false,
|
| 65 |
+
"single_word": false,
|
| 66 |
+
"special": true
|
| 67 |
+
},
|
| 68 |
+
"181895": {
|
| 69 |
+
"content": "<|reserved_special_token_4|>",
|
| 70 |
+
"lstrip": false,
|
| 71 |
+
"normalized": false,
|
| 72 |
+
"rstrip": false,
|
| 73 |
+
"single_word": false,
|
| 74 |
+
"special": true
|
| 75 |
+
},
|
| 76 |
+
"181896": {
|
| 77 |
+
"content": "<|eot_id|>",
|
| 78 |
+
"lstrip": false,
|
| 79 |
+
"normalized": false,
|
| 80 |
+
"rstrip": false,
|
| 81 |
+
"single_word": false,
|
| 82 |
+
"special": true
|
| 83 |
+
},
|
| 84 |
+
"181897": {
|
| 85 |
+
"content": "<|reserved_special_token_5|>",
|
| 86 |
+
"lstrip": false,
|
| 87 |
+
"normalized": false,
|
| 88 |
+
"rstrip": false,
|
| 89 |
+
"single_word": false,
|
| 90 |
+
"special": true
|
| 91 |
+
},
|
| 92 |
+
"181898": {
|
| 93 |
+
"content": "<|reserved_special_token_6|>",
|
| 94 |
+
"lstrip": false,
|
| 95 |
+
"normalized": false,
|
| 96 |
+
"rstrip": false,
|
| 97 |
+
"single_word": false,
|
| 98 |
+
"special": true
|
| 99 |
+
},
|
| 100 |
+
"181899": {
|
| 101 |
+
"content": "<|reserved_special_token_7|>",
|
| 102 |
+
"lstrip": false,
|
| 103 |
+
"normalized": false,
|
| 104 |
+
"rstrip": false,
|
| 105 |
+
"single_word": false,
|
| 106 |
+
"special": true
|
| 107 |
+
},
|
| 108 |
+
"181900": {
|
| 109 |
+
"content": "<|reserved_special_token_8|>",
|
| 110 |
+
"lstrip": false,
|
| 111 |
+
"normalized": false,
|
| 112 |
+
"rstrip": false,
|
| 113 |
+
"single_word": false,
|
| 114 |
+
"special": true
|
| 115 |
+
},
|
| 116 |
+
"181901": {
|
| 117 |
+
"content": "<|reserved_special_token_9|>",
|
| 118 |
+
"lstrip": false,
|
| 119 |
+
"normalized": false,
|
| 120 |
+
"rstrip": false,
|
| 121 |
+
"single_word": false,
|
| 122 |
+
"special": true
|
| 123 |
+
},
|
| 124 |
+
"181902": {
|
| 125 |
+
"content": "<|reserved_special_token_10|>",
|
| 126 |
+
"lstrip": false,
|
| 127 |
+
"normalized": false,
|
| 128 |
+
"rstrip": false,
|
| 129 |
+
"single_word": false,
|
| 130 |
+
"special": true
|
| 131 |
+
},
|
| 132 |
+
"181903": {
|
| 133 |
+
"content": "<|reserved_special_token_11|>",
|
| 134 |
+
"lstrip": false,
|
| 135 |
+
"normalized": false,
|
| 136 |
+
"rstrip": false,
|
| 137 |
+
"single_word": false,
|
| 138 |
+
"special": true
|
| 139 |
+
},
|
| 140 |
+
"181904": {
|
| 141 |
+
"content": "<|reserved_special_token_12|>",
|
| 142 |
+
"lstrip": false,
|
| 143 |
+
"normalized": false,
|
| 144 |
+
"rstrip": false,
|
| 145 |
+
"single_word": false,
|
| 146 |
+
"special": true
|
| 147 |
+
},
|
| 148 |
+
"181905": {
|
| 149 |
+
"content": "<|reserved_special_token_13|>",
|
| 150 |
+
"lstrip": false,
|
| 151 |
+
"normalized": false,
|
| 152 |
+
"rstrip": false,
|
| 153 |
+
"single_word": false,
|
| 154 |
+
"special": true
|
| 155 |
+
},
|
| 156 |
+
"181906": {
|
| 157 |
+
"content": "<|reserved_special_token_14|>",
|
| 158 |
+
"lstrip": false,
|
| 159 |
+
"normalized": false,
|
| 160 |
+
"rstrip": false,
|
| 161 |
+
"single_word": false,
|
| 162 |
+
"special": true
|
| 163 |
+
},
|
| 164 |
+
"181907": {
|
| 165 |
+
"content": "<|reserved_special_token_15|>",
|
| 166 |
+
"lstrip": false,
|
| 167 |
+
"normalized": false,
|
| 168 |
+
"rstrip": false,
|
| 169 |
+
"single_word": false,
|
| 170 |
+
"special": true
|
| 171 |
+
},
|
| 172 |
+
"181908": {
|
| 173 |
+
"content": "<|reserved_special_token_16|>",
|
| 174 |
+
"lstrip": false,
|
| 175 |
+
"normalized": false,
|
| 176 |
+
"rstrip": false,
|
| 177 |
+
"single_word": false,
|
| 178 |
+
"special": true
|
| 179 |
+
},
|
| 180 |
+
"181909": {
|
| 181 |
+
"content": "<|reserved_special_token_17|>",
|
| 182 |
+
"lstrip": false,
|
| 183 |
+
"normalized": false,
|
| 184 |
+
"rstrip": false,
|
| 185 |
+
"single_word": false,
|
| 186 |
+
"special": true
|
| 187 |
+
},
|
| 188 |
+
"181910": {
|
| 189 |
+
"content": "<|reserved_special_token_18|>",
|
| 190 |
+
"lstrip": false,
|
| 191 |
+
"normalized": false,
|
| 192 |
+
"rstrip": false,
|
| 193 |
+
"single_word": false,
|
| 194 |
+
"special": true
|
| 195 |
+
},
|
| 196 |
+
"181911": {
|
| 197 |
+
"content": "<|reserved_special_token_19|>",
|
| 198 |
+
"lstrip": false,
|
| 199 |
+
"normalized": false,
|
| 200 |
+
"rstrip": false,
|
| 201 |
+
"single_word": false,
|
| 202 |
+
"special": true
|
| 203 |
+
},
|
| 204 |
+
"181912": {
|
| 205 |
+
"content": "<|reserved_special_token_20|>",
|
| 206 |
+
"lstrip": false,
|
| 207 |
+
"normalized": false,
|
| 208 |
+
"rstrip": false,
|
| 209 |
+
"single_word": false,
|
| 210 |
+
"special": true
|
| 211 |
+
},
|
| 212 |
+
"181913": {
|
| 213 |
+
"content": "<|reserved_special_token_21|>",
|
| 214 |
+
"lstrip": false,
|
| 215 |
+
"normalized": false,
|
| 216 |
+
"rstrip": false,
|
| 217 |
+
"single_word": false,
|
| 218 |
+
"special": true
|
| 219 |
+
},
|
| 220 |
+
"181914": {
|
| 221 |
+
"content": "<|reserved_special_token_22|>",
|
| 222 |
+
"lstrip": false,
|
| 223 |
+
"normalized": false,
|
| 224 |
+
"rstrip": false,
|
| 225 |
+
"single_word": false,
|
| 226 |
+
"special": true
|
| 227 |
+
},
|
| 228 |
+
"181915": {
|
| 229 |
+
"content": "<|reserved_special_token_23|>",
|
| 230 |
+
"lstrip": false,
|
| 231 |
+
"normalized": false,
|
| 232 |
+
"rstrip": false,
|
| 233 |
+
"single_word": false,
|
| 234 |
+
"special": true
|
| 235 |
+
},
|
| 236 |
+
"181916": {
|
| 237 |
+
"content": "<|reserved_special_token_24|>",
|
| 238 |
+
"lstrip": false,
|
| 239 |
+
"normalized": false,
|
| 240 |
+
"rstrip": false,
|
| 241 |
+
"single_word": false,
|
| 242 |
+
"special": true
|
| 243 |
+
},
|
| 244 |
+
"181917": {
|
| 245 |
+
"content": "<|reserved_special_token_25|>",
|
| 246 |
+
"lstrip": false,
|
| 247 |
+
"normalized": false,
|
| 248 |
+
"rstrip": false,
|
| 249 |
+
"single_word": false,
|
| 250 |
+
"special": true
|
| 251 |
+
},
|
| 252 |
+
"181918": {
|
| 253 |
+
"content": "<|reserved_special_token_26|>",
|
| 254 |
+
"lstrip": false,
|
| 255 |
+
"normalized": false,
|
| 256 |
+
"rstrip": false,
|
| 257 |
+
"single_word": false,
|
| 258 |
+
"special": true
|
| 259 |
+
},
|
| 260 |
+
"181919": {
|
| 261 |
+
"content": "<|reserved_special_token_27|>",
|
| 262 |
+
"lstrip": false,
|
| 263 |
+
"normalized": false,
|
| 264 |
+
"rstrip": false,
|
| 265 |
+
"single_word": false,
|
| 266 |
+
"special": true
|
| 267 |
+
},
|
| 268 |
+
"181920": {
|
| 269 |
+
"content": "<|reserved_special_token_28|>",
|
| 270 |
+
"lstrip": false,
|
| 271 |
+
"normalized": false,
|
| 272 |
+
"rstrip": false,
|
| 273 |
+
"single_word": false,
|
| 274 |
+
"special": true
|
| 275 |
+
},
|
| 276 |
+
"181921": {
|
| 277 |
+
"content": "<|reserved_special_token_29|>",
|
| 278 |
+
"lstrip": false,
|
| 279 |
+
"normalized": false,
|
| 280 |
+
"rstrip": false,
|
| 281 |
+
"single_word": false,
|
| 282 |
+
"special": true
|
| 283 |
+
},
|
| 284 |
+
"181922": {
|
| 285 |
+
"content": "<|reserved_special_token_30|>",
|
| 286 |
+
"lstrip": false,
|
| 287 |
+
"normalized": false,
|
| 288 |
+
"rstrip": false,
|
| 289 |
+
"single_word": false,
|
| 290 |
+
"special": true
|
| 291 |
+
},
|
| 292 |
+
"181923": {
|
| 293 |
+
"content": "<|reserved_special_token_31|>",
|
| 294 |
+
"lstrip": false,
|
| 295 |
+
"normalized": false,
|
| 296 |
+
"rstrip": false,
|
| 297 |
+
"single_word": false,
|
| 298 |
+
"special": true
|
| 299 |
+
},
|
| 300 |
+
"181924": {
|
| 301 |
+
"content": "<|reserved_special_token_32|>",
|
| 302 |
+
"lstrip": false,
|
| 303 |
+
"normalized": false,
|
| 304 |
+
"rstrip": false,
|
| 305 |
+
"single_word": false,
|
| 306 |
+
"special": true
|
| 307 |
+
},
|
| 308 |
+
"181925": {
|
| 309 |
+
"content": "<|reserved_special_token_33|>",
|
| 310 |
+
"lstrip": false,
|
| 311 |
+
"normalized": false,
|
| 312 |
+
"rstrip": false,
|
| 313 |
+
"single_word": false,
|
| 314 |
+
"special": true
|
| 315 |
+
},
|
| 316 |
+
"181926": {
|
| 317 |
+
"content": "<|reserved_special_token_34|>",
|
| 318 |
+
"lstrip": false,
|
| 319 |
+
"normalized": false,
|
| 320 |
+
"rstrip": false,
|
| 321 |
+
"single_word": false,
|
| 322 |
+
"special": true
|
| 323 |
+
},
|
| 324 |
+
"181927": {
|
| 325 |
+
"content": "<|reserved_special_token_35|>",
|
| 326 |
+
"lstrip": false,
|
| 327 |
+
"normalized": false,
|
| 328 |
+
"rstrip": false,
|
| 329 |
+
"single_word": false,
|
| 330 |
+
"special": true
|
| 331 |
+
},
|
| 332 |
+
"181928": {
|
| 333 |
+
"content": "<|reserved_special_token_36|>",
|
| 334 |
+
"lstrip": false,
|
| 335 |
+
"normalized": false,
|
| 336 |
+
"rstrip": false,
|
| 337 |
+
"single_word": false,
|
| 338 |
+
"special": true
|
| 339 |
+
},
|
| 340 |
+
"181929": {
|
| 341 |
+
"content": "<|reserved_special_token_37|>",
|
| 342 |
+
"lstrip": false,
|
| 343 |
+
"normalized": false,
|
| 344 |
+
"rstrip": false,
|
| 345 |
+
"single_word": false,
|
| 346 |
+
"special": true
|
| 347 |
+
},
|
| 348 |
+
"181930": {
|
| 349 |
+
"content": "<|reserved_special_token_38|>",
|
| 350 |
+
"lstrip": false,
|
| 351 |
+
"normalized": false,
|
| 352 |
+
"rstrip": false,
|
| 353 |
+
"single_word": false,
|
| 354 |
+
"special": true
|
| 355 |
+
},
|
| 356 |
+
"181931": {
|
| 357 |
+
"content": "<|reserved_special_token_39|>",
|
| 358 |
+
"lstrip": false,
|
| 359 |
+
"normalized": false,
|
| 360 |
+
"rstrip": false,
|
| 361 |
+
"single_word": false,
|
| 362 |
+
"special": true
|
| 363 |
+
},
|
| 364 |
+
"181932": {
|
| 365 |
+
"content": "<|reserved_special_token_40|>",
|
| 366 |
+
"lstrip": false,
|
| 367 |
+
"normalized": false,
|
| 368 |
+
"rstrip": false,
|
| 369 |
+
"single_word": false,
|
| 370 |
+
"special": true
|
| 371 |
+
},
|
| 372 |
+
"181933": {
|
| 373 |
+
"content": "<|reserved_special_token_41|>",
|
| 374 |
+
"lstrip": false,
|
| 375 |
+
"normalized": false,
|
| 376 |
+
"rstrip": false,
|
| 377 |
+
"single_word": false,
|
| 378 |
+
"special": true
|
| 379 |
+
},
|
| 380 |
+
"181934": {
|
| 381 |
+
"content": "<|reserved_special_token_42|>",
|
| 382 |
+
"lstrip": false,
|
| 383 |
+
"normalized": false,
|
| 384 |
+
"rstrip": false,
|
| 385 |
+
"single_word": false,
|
| 386 |
+
"special": true
|
| 387 |
+
},
|
| 388 |
+
"181935": {
|
| 389 |
+
"content": "<|reserved_special_token_43|>",
|
| 390 |
+
"lstrip": false,
|
| 391 |
+
"normalized": false,
|
| 392 |
+
"rstrip": false,
|
| 393 |
+
"single_word": false,
|
| 394 |
+
"special": true
|
| 395 |
+
},
|
| 396 |
+
"181936": {
|
| 397 |
+
"content": "<|reserved_special_token_44|>",
|
| 398 |
+
"lstrip": false,
|
| 399 |
+
"normalized": false,
|
| 400 |
+
"rstrip": false,
|
| 401 |
+
"single_word": false,
|
| 402 |
+
"special": true
|
| 403 |
+
},
|
| 404 |
+
"181937": {
|
| 405 |
+
"content": "<|reserved_special_token_45|>",
|
| 406 |
+
"lstrip": false,
|
| 407 |
+
"normalized": false,
|
| 408 |
+
"rstrip": false,
|
| 409 |
+
"single_word": false,
|
| 410 |
+
"special": true
|
| 411 |
+
},
|
| 412 |
+
"181938": {
|
| 413 |
+
"content": "<|reserved_special_token_46|>",
|
| 414 |
+
"lstrip": false,
|
| 415 |
+
"normalized": false,
|
| 416 |
+
"rstrip": false,
|
| 417 |
+
"single_word": false,
|
| 418 |
+
"special": true
|
| 419 |
+
},
|
| 420 |
+
"181939": {
|
| 421 |
+
"content": "<|reserved_special_token_47|>",
|
| 422 |
+
"lstrip": false,
|
| 423 |
+
"normalized": false,
|
| 424 |
+
"rstrip": false,
|
| 425 |
+
"single_word": false,
|
| 426 |
+
"special": true
|
| 427 |
+
},
|
| 428 |
+
"181940": {
|
| 429 |
+
"content": "<|reserved_special_token_48|>",
|
| 430 |
+
"lstrip": false,
|
| 431 |
+
"normalized": false,
|
| 432 |
+
"rstrip": false,
|
| 433 |
+
"single_word": false,
|
| 434 |
+
"special": true
|
| 435 |
+
},
|
| 436 |
+
"181941": {
|
| 437 |
+
"content": "<|reserved_special_token_49|>",
|
| 438 |
+
"lstrip": false,
|
| 439 |
+
"normalized": false,
|
| 440 |
+
"rstrip": false,
|
| 441 |
+
"single_word": false,
|
| 442 |
+
"special": true
|
| 443 |
+
},
|
| 444 |
+
"181942": {
|
| 445 |
+
"content": "<|reserved_special_token_50|>",
|
| 446 |
+
"lstrip": false,
|
| 447 |
+
"normalized": false,
|
| 448 |
+
"rstrip": false,
|
| 449 |
+
"single_word": false,
|
| 450 |
+
"special": true
|
| 451 |
+
},
|
| 452 |
+
"181943": {
|
| 453 |
+
"content": "<|reserved_special_token_51|>",
|
| 454 |
+
"lstrip": false,
|
| 455 |
+
"normalized": false,
|
| 456 |
+
"rstrip": false,
|
| 457 |
+
"single_word": false,
|
| 458 |
+
"special": true
|
| 459 |
+
},
|
| 460 |
+
"181944": {
|
| 461 |
+
"content": "<|reserved_special_token_52|>",
|
| 462 |
+
"lstrip": false,
|
| 463 |
+
"normalized": false,
|
| 464 |
+
"rstrip": false,
|
| 465 |
+
"single_word": false,
|
| 466 |
+
"special": true
|
| 467 |
+
},
|
| 468 |
+
"181945": {
|
| 469 |
+
"content": "<|reserved_special_token_53|>",
|
| 470 |
+
"lstrip": false,
|
| 471 |
+
"normalized": false,
|
| 472 |
+
"rstrip": false,
|
| 473 |
+
"single_word": false,
|
| 474 |
+
"special": true
|
| 475 |
+
},
|
| 476 |
+
"181946": {
|
| 477 |
+
"content": "<|reserved_special_token_54|>",
|
| 478 |
+
"lstrip": false,
|
| 479 |
+
"normalized": false,
|
| 480 |
+
"rstrip": false,
|
| 481 |
+
"single_word": false,
|
| 482 |
+
"special": true
|
| 483 |
+
},
|
| 484 |
+
"181947": {
|
| 485 |
+
"content": "<|reserved_special_token_55|>",
|
| 486 |
+
"lstrip": false,
|
| 487 |
+
"normalized": false,
|
| 488 |
+
"rstrip": false,
|
| 489 |
+
"single_word": false,
|
| 490 |
+
"special": true
|
| 491 |
+
},
|
| 492 |
+
"181948": {
|
| 493 |
+
"content": "<|reserved_special_token_56|>",
|
| 494 |
+
"lstrip": false,
|
| 495 |
+
"normalized": false,
|
| 496 |
+
"rstrip": false,
|
| 497 |
+
"single_word": false,
|
| 498 |
+
"special": true
|
| 499 |
+
},
|
| 500 |
+
"181949": {
|
| 501 |
+
"content": "<|reserved_special_token_57|>",
|
| 502 |
+
"lstrip": false,
|
| 503 |
+
"normalized": false,
|
| 504 |
+
"rstrip": false,
|
| 505 |
+
"single_word": false,
|
| 506 |
+
"special": true
|
| 507 |
+
},
|
| 508 |
+
"181950": {
|
| 509 |
+
"content": "<|reserved_special_token_58|>",
|
| 510 |
+
"lstrip": false,
|
| 511 |
+
"normalized": false,
|
| 512 |
+
"rstrip": false,
|
| 513 |
+
"single_word": false,
|
| 514 |
+
"special": true
|
| 515 |
+
},
|
| 516 |
+
"181951": {
|
| 517 |
+
"content": "<|reserved_special_token_59|>",
|
| 518 |
+
"lstrip": false,
|
| 519 |
+
"normalized": false,
|
| 520 |
+
"rstrip": false,
|
| 521 |
+
"single_word": false,
|
| 522 |
+
"special": true
|
| 523 |
+
},
|
| 524 |
+
"181952": {
|
| 525 |
+
"content": "<|reserved_special_token_60|>",
|
| 526 |
+
"lstrip": false,
|
| 527 |
+
"normalized": false,
|
| 528 |
+
"rstrip": false,
|
| 529 |
+
"single_word": false,
|
| 530 |
+
"special": true
|
| 531 |
+
},
|
| 532 |
+
"181953": {
|
| 533 |
+
"content": "<|reserved_special_token_61|>",
|
| 534 |
+
"lstrip": false,
|
| 535 |
+
"normalized": false,
|
| 536 |
+
"rstrip": false,
|
| 537 |
+
"single_word": false,
|
| 538 |
+
"special": true
|
| 539 |
+
},
|
| 540 |
+
"181954": {
|
| 541 |
+
"content": "<|reserved_special_token_62|>",
|
| 542 |
+
"lstrip": false,
|
| 543 |
+
"normalized": false,
|
| 544 |
+
"rstrip": false,
|
| 545 |
+
"single_word": false,
|
| 546 |
+
"special": true
|
| 547 |
+
},
|
| 548 |
+
"181955": {
|
| 549 |
+
"content": "<|reserved_special_token_63|>",
|
| 550 |
+
"lstrip": false,
|
| 551 |
+
"normalized": false,
|
| 552 |
+
"rstrip": false,
|
| 553 |
+
"single_word": false,
|
| 554 |
+
"special": true
|
| 555 |
+
},
|
| 556 |
+
"181956": {
|
| 557 |
+
"content": "<|reserved_special_token_64|>",
|
| 558 |
+
"lstrip": false,
|
| 559 |
+
"normalized": false,
|
| 560 |
+
"rstrip": false,
|
| 561 |
+
"single_word": false,
|
| 562 |
+
"special": true
|
| 563 |
+
},
|
| 564 |
+
"181957": {
|
| 565 |
+
"content": "<|reserved_special_token_65|>",
|
| 566 |
+
"lstrip": false,
|
| 567 |
+
"normalized": false,
|
| 568 |
+
"rstrip": false,
|
| 569 |
+
"single_word": false,
|
| 570 |
+
"special": true
|
| 571 |
+
},
|
| 572 |
+
"181958": {
|
| 573 |
+
"content": "<|reserved_special_token_66|>",
|
| 574 |
+
"lstrip": false,
|
| 575 |
+
"normalized": false,
|
| 576 |
+
"rstrip": false,
|
| 577 |
+
"single_word": false,
|
| 578 |
+
"special": true
|
| 579 |
+
},
|
| 580 |
+
"181959": {
|
| 581 |
+
"content": "<|reserved_special_token_67|>",
|
| 582 |
+
"lstrip": false,
|
| 583 |
+
"normalized": false,
|
| 584 |
+
"rstrip": false,
|
| 585 |
+
"single_word": false,
|
| 586 |
+
"special": true
|
| 587 |
+
},
|
| 588 |
+
"181960": {
|
| 589 |
+
"content": "<|reserved_special_token_68|>",
|
| 590 |
+
"lstrip": false,
|
| 591 |
+
"normalized": false,
|
| 592 |
+
"rstrip": false,
|
| 593 |
+
"single_word": false,
|
| 594 |
+
"special": true
|
| 595 |
+
},
|
| 596 |
+
"181961": {
|
| 597 |
+
"content": "<|reserved_special_token_69|>",
|
| 598 |
+
"lstrip": false,
|
| 599 |
+
"normalized": false,
|
| 600 |
+
"rstrip": false,
|
| 601 |
+
"single_word": false,
|
| 602 |
+
"special": true
|
| 603 |
+
},
|
| 604 |
+
"181962": {
|
| 605 |
+
"content": "<|reserved_special_token_70|>",
|
| 606 |
+
"lstrip": false,
|
| 607 |
+
"normalized": false,
|
| 608 |
+
"rstrip": false,
|
| 609 |
+
"single_word": false,
|
| 610 |
+
"special": true
|
| 611 |
+
},
|
| 612 |
+
"181963": {
|
| 613 |
+
"content": "<|reserved_special_token_71|>",
|
| 614 |
+
"lstrip": false,
|
| 615 |
+
"normalized": false,
|
| 616 |
+
"rstrip": false,
|
| 617 |
+
"single_word": false,
|
| 618 |
+
"special": true
|
| 619 |
+
},
|
| 620 |
+
"181964": {
|
| 621 |
+
"content": "<|reserved_special_token_72|>",
|
| 622 |
+
"lstrip": false,
|
| 623 |
+
"normalized": false,
|
| 624 |
+
"rstrip": false,
|
| 625 |
+
"single_word": false,
|
| 626 |
+
"special": true
|
| 627 |
+
},
|
| 628 |
+
"181965": {
|
| 629 |
+
"content": "<|reserved_special_token_73|>",
|
| 630 |
+
"lstrip": false,
|
| 631 |
+
"normalized": false,
|
| 632 |
+
"rstrip": false,
|
| 633 |
+
"single_word": false,
|
| 634 |
+
"special": true
|
| 635 |
+
},
|
| 636 |
+
"181966": {
|
| 637 |
+
"content": "<|reserved_special_token_74|>",
|
| 638 |
+
"lstrip": false,
|
| 639 |
+
"normalized": false,
|
| 640 |
+
"rstrip": false,
|
| 641 |
+
"single_word": false,
|
| 642 |
+
"special": true
|
| 643 |
+
},
|
| 644 |
+
"181967": {
|
| 645 |
+
"content": "<|reserved_special_token_75|>",
|
| 646 |
+
"lstrip": false,
|
| 647 |
+
"normalized": false,
|
| 648 |
+
"rstrip": false,
|
| 649 |
+
"single_word": false,
|
| 650 |
+
"special": true
|
| 651 |
+
},
|
| 652 |
+
"181968": {
|
| 653 |
+
"content": "<|reserved_special_token_76|>",
|
| 654 |
+
"lstrip": false,
|
| 655 |
+
"normalized": false,
|
| 656 |
+
"rstrip": false,
|
| 657 |
+
"single_word": false,
|
| 658 |
+
"special": true
|
| 659 |
+
},
|
| 660 |
+
"181969": {
|
| 661 |
+
"content": "<|reserved_special_token_77|>",
|
| 662 |
+
"lstrip": false,
|
| 663 |
+
"normalized": false,
|
| 664 |
+
"rstrip": false,
|
| 665 |
+
"single_word": false,
|
| 666 |
+
"special": true
|
| 667 |
+
},
|
| 668 |
+
"181970": {
|
| 669 |
+
"content": "<|reserved_special_token_78|>",
|
| 670 |
+
"lstrip": false,
|
| 671 |
+
"normalized": false,
|
| 672 |
+
"rstrip": false,
|
| 673 |
+
"single_word": false,
|
| 674 |
+
"special": true
|
| 675 |
+
},
|
| 676 |
+
"181971": {
|
| 677 |
+
"content": "<|reserved_special_token_79|>",
|
| 678 |
+
"lstrip": false,
|
| 679 |
+
"normalized": false,
|
| 680 |
+
"rstrip": false,
|
| 681 |
+
"single_word": false,
|
| 682 |
+
"special": true
|
| 683 |
+
},
|
| 684 |
+
"181972": {
|
| 685 |
+
"content": "<|reserved_special_token_80|>",
|
| 686 |
+
"lstrip": false,
|
| 687 |
+
"normalized": false,
|
| 688 |
+
"rstrip": false,
|
| 689 |
+
"single_word": false,
|
| 690 |
+
"special": true
|
| 691 |
+
},
|
| 692 |
+
"181973": {
|
| 693 |
+
"content": "<|reserved_special_token_81|>",
|
| 694 |
+
"lstrip": false,
|
| 695 |
+
"normalized": false,
|
| 696 |
+
"rstrip": false,
|
| 697 |
+
"single_word": false,
|
| 698 |
+
"special": true
|
| 699 |
+
},
|
| 700 |
+
"181974": {
|
| 701 |
+
"content": "<|reserved_special_token_82|>",
|
| 702 |
+
"lstrip": false,
|
| 703 |
+
"normalized": false,
|
| 704 |
+
"rstrip": false,
|
| 705 |
+
"single_word": false,
|
| 706 |
+
"special": true
|
| 707 |
+
},
|
| 708 |
+
"181975": {
|
| 709 |
+
"content": "<|reserved_special_token_83|>",
|
| 710 |
+
"lstrip": false,
|
| 711 |
+
"normalized": false,
|
| 712 |
+
"rstrip": false,
|
| 713 |
+
"single_word": false,
|
| 714 |
+
"special": true
|
| 715 |
+
},
|
| 716 |
+
"181976": {
|
| 717 |
+
"content": "<|reserved_special_token_84|>",
|
| 718 |
+
"lstrip": false,
|
| 719 |
+
"normalized": false,
|
| 720 |
+
"rstrip": false,
|
| 721 |
+
"single_word": false,
|
| 722 |
+
"special": true
|
| 723 |
+
},
|
| 724 |
+
"181977": {
|
| 725 |
+
"content": "<|reserved_special_token_85|>",
|
| 726 |
+
"lstrip": false,
|
| 727 |
+
"normalized": false,
|
| 728 |
+
"rstrip": false,
|
| 729 |
+
"single_word": false,
|
| 730 |
+
"special": true
|
| 731 |
+
},
|
| 732 |
+
"181978": {
|
| 733 |
+
"content": "<|reserved_special_token_86|>",
|
| 734 |
+
"lstrip": false,
|
| 735 |
+
"normalized": false,
|
| 736 |
+
"rstrip": false,
|
| 737 |
+
"single_word": false,
|
| 738 |
+
"special": true
|
| 739 |
+
},
|
| 740 |
+
"181979": {
|
| 741 |
+
"content": "<|reserved_special_token_87|>",
|
| 742 |
+
"lstrip": false,
|
| 743 |
+
"normalized": false,
|
| 744 |
+
"rstrip": false,
|
| 745 |
+
"single_word": false,
|
| 746 |
+
"special": true
|
| 747 |
+
},
|
| 748 |
+
"181980": {
|
| 749 |
+
"content": "<|reserved_special_token_88|>",
|
| 750 |
+
"lstrip": false,
|
| 751 |
+
"normalized": false,
|
| 752 |
+
"rstrip": false,
|
| 753 |
+
"single_word": false,
|
| 754 |
+
"special": true
|
| 755 |
+
},
|
| 756 |
+
"181981": {
|
| 757 |
+
"content": "<|reserved_special_token_89|>",
|
| 758 |
+
"lstrip": false,
|
| 759 |
+
"normalized": false,
|
| 760 |
+
"rstrip": false,
|
| 761 |
+
"single_word": false,
|
| 762 |
+
"special": true
|
| 763 |
+
},
|
| 764 |
+
"181982": {
|
| 765 |
+
"content": "<|reserved_special_token_90|>",
|
| 766 |
+
"lstrip": false,
|
| 767 |
+
"normalized": false,
|
| 768 |
+
"rstrip": false,
|
| 769 |
+
"single_word": false,
|
| 770 |
+
"special": true
|
| 771 |
+
},
|
| 772 |
+
"181983": {
|
| 773 |
+
"content": "<|reserved_special_token_91|>",
|
| 774 |
+
"lstrip": false,
|
| 775 |
+
"normalized": false,
|
| 776 |
+
"rstrip": false,
|
| 777 |
+
"single_word": false,
|
| 778 |
+
"special": true
|
| 779 |
+
},
|
| 780 |
+
"181984": {
|
| 781 |
+
"content": "<|reserved_special_token_92|>",
|
| 782 |
+
"lstrip": false,
|
| 783 |
+
"normalized": false,
|
| 784 |
+
"rstrip": false,
|
| 785 |
+
"single_word": false,
|
| 786 |
+
"special": true
|
| 787 |
+
},
|
| 788 |
+
"181985": {
|
| 789 |
+
"content": "<|reserved_special_token_93|>",
|
| 790 |
+
"lstrip": false,
|
| 791 |
+
"normalized": false,
|
| 792 |
+
"rstrip": false,
|
| 793 |
+
"single_word": false,
|
| 794 |
+
"special": true
|
| 795 |
+
},
|
| 796 |
+
"181986": {
|
| 797 |
+
"content": "<|reserved_special_token_94|>",
|
| 798 |
+
"lstrip": false,
|
| 799 |
+
"normalized": false,
|
| 800 |
+
"rstrip": false,
|
| 801 |
+
"single_word": false,
|
| 802 |
+
"special": true
|
| 803 |
+
},
|
| 804 |
+
"181987": {
|
| 805 |
+
"content": "<|reserved_special_token_95|>",
|
| 806 |
+
"lstrip": false,
|
| 807 |
+
"normalized": false,
|
| 808 |
+
"rstrip": false,
|
| 809 |
+
"single_word": false,
|
| 810 |
+
"special": true
|
| 811 |
+
},
|
| 812 |
+
"181988": {
|
| 813 |
+
"content": "<|reserved_special_token_96|>",
|
| 814 |
+
"lstrip": false,
|
| 815 |
+
"normalized": false,
|
| 816 |
+
"rstrip": false,
|
| 817 |
+
"single_word": false,
|
| 818 |
+
"special": true
|
| 819 |
+
},
|
| 820 |
+
"181989": {
|
| 821 |
+
"content": "<|reserved_special_token_97|>",
|
| 822 |
+
"lstrip": false,
|
| 823 |
+
"normalized": false,
|
| 824 |
+
"rstrip": false,
|
| 825 |
+
"single_word": false,
|
| 826 |
+
"special": true
|
| 827 |
+
},
|
| 828 |
+
"181990": {
|
| 829 |
+
"content": "<|reserved_special_token_98|>",
|
| 830 |
+
"lstrip": false,
|
| 831 |
+
"normalized": false,
|
| 832 |
+
"rstrip": false,
|
| 833 |
+
"single_word": false,
|
| 834 |
+
"special": true
|
| 835 |
+
},
|
| 836 |
+
"181991": {
|
| 837 |
+
"content": "<|reserved_special_token_99|>",
|
| 838 |
+
"lstrip": false,
|
| 839 |
+
"normalized": false,
|
| 840 |
+
"rstrip": false,
|
| 841 |
+
"single_word": false,
|
| 842 |
+
"special": true
|
| 843 |
+
},
|
| 844 |
+
"181992": {
|
| 845 |
+
"content": "<|reserved_special_token_100|>",
|
| 846 |
+
"lstrip": false,
|
| 847 |
+
"normalized": false,
|
| 848 |
+
"rstrip": false,
|
| 849 |
+
"single_word": false,
|
| 850 |
+
"special": true
|
| 851 |
+
},
|
| 852 |
+
"181993": {
|
| 853 |
+
"content": "<|reserved_special_token_101|>",
|
| 854 |
+
"lstrip": false,
|
| 855 |
+
"normalized": false,
|
| 856 |
+
"rstrip": false,
|
| 857 |
+
"single_word": false,
|
| 858 |
+
"special": true
|
| 859 |
+
},
|
| 860 |
+
"181994": {
|
| 861 |
+
"content": "<|reserved_special_token_102|>",
|
| 862 |
+
"lstrip": false,
|
| 863 |
+
"normalized": false,
|
| 864 |
+
"rstrip": false,
|
| 865 |
+
"single_word": false,
|
| 866 |
+
"special": true
|
| 867 |
+
},
|
| 868 |
+
"181995": {
|
| 869 |
+
"content": "<|reserved_special_token_103|>",
|
| 870 |
+
"lstrip": false,
|
| 871 |
+
"normalized": false,
|
| 872 |
+
"rstrip": false,
|
| 873 |
+
"single_word": false,
|
| 874 |
+
"special": true
|
| 875 |
+
},
|
| 876 |
+
"181996": {
|
| 877 |
+
"content": "<|reserved_special_token_104|>",
|
| 878 |
+
"lstrip": false,
|
| 879 |
+
"normalized": false,
|
| 880 |
+
"rstrip": false,
|
| 881 |
+
"single_word": false,
|
| 882 |
+
"special": true
|
| 883 |
+
},
|
| 884 |
+
"181997": {
|
| 885 |
+
"content": "<|reserved_special_token_105|>",
|
| 886 |
+
"lstrip": false,
|
| 887 |
+
"normalized": false,
|
| 888 |
+
"rstrip": false,
|
| 889 |
+
"single_word": false,
|
| 890 |
+
"special": true
|
| 891 |
+
},
|
| 892 |
+
"181998": {
|
| 893 |
+
"content": "<|reserved_special_token_106|>",
|
| 894 |
+
"lstrip": false,
|
| 895 |
+
"normalized": false,
|
| 896 |
+
"rstrip": false,
|
| 897 |
+
"single_word": false,
|
| 898 |
+
"special": true
|
| 899 |
+
},
|
| 900 |
+
"181999": {
|
| 901 |
+
"content": "<|reserved_special_token_107|>",
|
| 902 |
+
"lstrip": false,
|
| 903 |
+
"normalized": false,
|
| 904 |
+
"rstrip": false,
|
| 905 |
+
"single_word": false,
|
| 906 |
+
"special": true
|
| 907 |
+
},
|
| 908 |
+
"182000": {
|
| 909 |
+
"content": "<|reserved_special_token_108|>",
|
| 910 |
+
"lstrip": false,
|
| 911 |
+
"normalized": false,
|
| 912 |
+
"rstrip": false,
|
| 913 |
+
"single_word": false,
|
| 914 |
+
"special": true
|
| 915 |
+
},
|
| 916 |
+
"182001": {
|
| 917 |
+
"content": "<|reserved_special_token_109|>",
|
| 918 |
+
"lstrip": false,
|
| 919 |
+
"normalized": false,
|
| 920 |
+
"rstrip": false,
|
| 921 |
+
"single_word": false,
|
| 922 |
+
"special": true
|
| 923 |
+
},
|
| 924 |
+
"182002": {
|
| 925 |
+
"content": "<|reserved_special_token_110|>",
|
| 926 |
+
"lstrip": false,
|
| 927 |
+
"normalized": false,
|
| 928 |
+
"rstrip": false,
|
| 929 |
+
"single_word": false,
|
| 930 |
+
"special": true
|
| 931 |
+
},
|
| 932 |
+
"182003": {
|
| 933 |
+
"content": "<|reserved_special_token_111|>",
|
| 934 |
+
"lstrip": false,
|
| 935 |
+
"normalized": false,
|
| 936 |
+
"rstrip": false,
|
| 937 |
+
"single_word": false,
|
| 938 |
+
"special": true
|
| 939 |
+
},
|
| 940 |
+
"182004": {
|
| 941 |
+
"content": "<|reserved_special_token_112|>",
|
| 942 |
+
"lstrip": false,
|
| 943 |
+
"normalized": false,
|
| 944 |
+
"rstrip": false,
|
| 945 |
+
"single_word": false,
|
| 946 |
+
"special": true
|
| 947 |
+
},
|
| 948 |
+
"182005": {
|
| 949 |
+
"content": "<|reserved_special_token_113|>",
|
| 950 |
+
"lstrip": false,
|
| 951 |
+
"normalized": false,
|
| 952 |
+
"rstrip": false,
|
| 953 |
+
"single_word": false,
|
| 954 |
+
"special": true
|
| 955 |
+
},
|
| 956 |
+
"182006": {
|
| 957 |
+
"content": "<|reserved_special_token_114|>",
|
| 958 |
+
"lstrip": false,
|
| 959 |
+
"normalized": false,
|
| 960 |
+
"rstrip": false,
|
| 961 |
+
"single_word": false,
|
| 962 |
+
"special": true
|
| 963 |
+
},
|
| 964 |
+
"182007": {
|
| 965 |
+
"content": "<|reserved_special_token_115|>",
|
| 966 |
+
"lstrip": false,
|
| 967 |
+
"normalized": false,
|
| 968 |
+
"rstrip": false,
|
| 969 |
+
"single_word": false,
|
| 970 |
+
"special": true
|
| 971 |
+
},
|
| 972 |
+
"182008": {
|
| 973 |
+
"content": "<|reserved_special_token_116|>",
|
| 974 |
+
"lstrip": false,
|
| 975 |
+
"normalized": false,
|
| 976 |
+
"rstrip": false,
|
| 977 |
+
"single_word": false,
|
| 978 |
+
"special": true
|
| 979 |
+
},
|
| 980 |
+
"182009": {
|
| 981 |
+
"content": "<|reserved_special_token_117|>",
|
| 982 |
+
"lstrip": false,
|
| 983 |
+
"normalized": false,
|
| 984 |
+
"rstrip": false,
|
| 985 |
+
"single_word": false,
|
| 986 |
+
"special": true
|
| 987 |
+
},
|
| 988 |
+
"182010": {
|
| 989 |
+
"content": "<|reserved_special_token_118|>",
|
| 990 |
+
"lstrip": false,
|
| 991 |
+
"normalized": false,
|
| 992 |
+
"rstrip": false,
|
| 993 |
+
"single_word": false,
|
| 994 |
+
"special": true
|
| 995 |
+
},
|
| 996 |
+
"182011": {
|
| 997 |
+
"content": "<|reserved_special_token_119|>",
|
| 998 |
+
"lstrip": false,
|
| 999 |
+
"normalized": false,
|
| 1000 |
+
"rstrip": false,
|
| 1001 |
+
"single_word": false,
|
| 1002 |
+
"special": true
|
| 1003 |
+
},
|
| 1004 |
+
"182012": {
|
| 1005 |
+
"content": "<|reserved_special_token_120|>",
|
| 1006 |
+
"lstrip": false,
|
| 1007 |
+
"normalized": false,
|
| 1008 |
+
"rstrip": false,
|
| 1009 |
+
"single_word": false,
|
| 1010 |
+
"special": true
|
| 1011 |
+
},
|
| 1012 |
+
"182013": {
|
| 1013 |
+
"content": "<|reserved_special_token_121|>",
|
| 1014 |
+
"lstrip": false,
|
| 1015 |
+
"normalized": false,
|
| 1016 |
+
"rstrip": false,
|
| 1017 |
+
"single_word": false,
|
| 1018 |
+
"special": true
|
| 1019 |
+
},
|
| 1020 |
+
"182014": {
|
| 1021 |
+
"content": "<|reserved_special_token_122|>",
|
| 1022 |
+
"lstrip": false,
|
| 1023 |
+
"normalized": false,
|
| 1024 |
+
"rstrip": false,
|
| 1025 |
+
"single_word": false,
|
| 1026 |
+
"special": true
|
| 1027 |
+
},
|
| 1028 |
+
"182015": {
|
| 1029 |
+
"content": "<|reserved_special_token_123|>",
|
| 1030 |
+
"lstrip": false,
|
| 1031 |
+
"normalized": false,
|
| 1032 |
+
"rstrip": false,
|
| 1033 |
+
"single_word": false,
|
| 1034 |
+
"special": true
|
| 1035 |
+
},
|
| 1036 |
+
"182016": {
|
| 1037 |
+
"content": "<img>",
|
| 1038 |
+
"lstrip": false,
|
| 1039 |
+
"normalized": false,
|
| 1040 |
+
"rstrip": false,
|
| 1041 |
+
"single_word": false,
|
| 1042 |
+
"special": true
|
| 1043 |
+
},
|
| 1044 |
+
"182017": {
|
| 1045 |
+
"content": "</img>",
|
| 1046 |
+
"lstrip": false,
|
| 1047 |
+
"normalized": false,
|
| 1048 |
+
"rstrip": false,
|
| 1049 |
+
"single_word": false,
|
| 1050 |
+
"special": true
|
| 1051 |
+
},
|
| 1052 |
+
"182018": {
|
| 1053 |
+
"content": "<IMG_CONTEXT>",
|
| 1054 |
+
"lstrip": false,
|
| 1055 |
+
"normalized": false,
|
| 1056 |
+
"rstrip": false,
|
| 1057 |
+
"single_word": false,
|
| 1058 |
+
"special": true
|
| 1059 |
+
},
|
| 1060 |
+
"182019": {
|
| 1061 |
+
"content": "<quad>",
|
| 1062 |
+
"lstrip": false,
|
| 1063 |
+
"normalized": false,
|
| 1064 |
+
"rstrip": false,
|
| 1065 |
+
"single_word": false,
|
| 1066 |
+
"special": true
|
| 1067 |
+
},
|
| 1068 |
+
"182020": {
|
| 1069 |
+
"content": "</quad>",
|
| 1070 |
+
"lstrip": false,
|
| 1071 |
+
"normalized": false,
|
| 1072 |
+
"rstrip": false,
|
| 1073 |
+
"single_word": false,
|
| 1074 |
+
"special": true
|
| 1075 |
+
},
|
| 1076 |
+
"182021": {
|
| 1077 |
+
"content": "<ref>",
|
| 1078 |
+
"lstrip": false,
|
| 1079 |
+
"normalized": false,
|
| 1080 |
+
"rstrip": false,
|
| 1081 |
+
"single_word": false,
|
| 1082 |
+
"special": true
|
| 1083 |
+
},
|
| 1084 |
+
"182022": {
|
| 1085 |
+
"content": "</ref>",
|
| 1086 |
+
"lstrip": false,
|
| 1087 |
+
"normalized": false,
|
| 1088 |
+
"rstrip": false,
|
| 1089 |
+
"single_word": false,
|
| 1090 |
+
"special": true
|
| 1091 |
+
},
|
| 1092 |
+
"182023": {
|
| 1093 |
+
"content": "<box>",
|
| 1094 |
+
"lstrip": false,
|
| 1095 |
+
"normalized": false,
|
| 1096 |
+
"rstrip": false,
|
| 1097 |
+
"single_word": false,
|
| 1098 |
+
"special": true
|
| 1099 |
+
},
|
| 1100 |
+
"182024": {
|
| 1101 |
+
"content": "</box>",
|
| 1102 |
+
"lstrip": false,
|
| 1103 |
+
"normalized": false,
|
| 1104 |
+
"rstrip": false,
|
| 1105 |
+
"single_word": false,
|
| 1106 |
+
"special": true
|
| 1107 |
+
}
|
| 1108 |
+
},
|
| 1109 |
+
"bos_token": "<|begin_of_text|>",
|
| 1110 |
+
"chat_template": "{% set loop_messages = messages %}{% if loop_messages[0]['role'] != 'system' %}{% set loop_messages = [{'role': 'system', 'content': '你是Qianfan-VL，由百度智能云千帆团队研发的多模态大模型。'}] + loop_messages %}{% endif %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n' + message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}",
|
| 1111 |
+
"clean_up_tokenization_spaces": true,
|
| 1112 |
+
"eos_token": "<|eot_id|>",
|
| 1113 |
+
"model_input_names": [
|
| 1114 |
+
"input_ids",
|
| 1115 |
+
"attention_mask"
|
| 1116 |
+
],
|
| 1117 |
+
"model_max_length": 32768,
|
| 1118 |
+
"pad_token": "<|end_of_text|>",
|
| 1119 |
+
"tokenizer_class": "PreTrainedTokenizerFast"
|
| 1120 |
+
}
|