File size: 5,270 Bytes
e9bf2b4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 |
---
library_name: transformers
license: apache-2.0
base_model: mistralai/Mistral-Nemo-Instruct-2407
tags:
- axolotl
- generated_from_trainer
<!-- datasets:
- ./train_data.jsonl-->
model-index:
- name: linabot
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
<details><summary>See axolotl config</summary>
axolotl version: `0.8.0`
```yaml
base_model: mistralai/Mistral-Nemo-Instruct-2407
model_type: MistralForCausalLM
hub_model_id: Alignment-Lab-AI/linabot
strict: false
chat_template: tokenizer_default
plugins:
- axolotl.integrations.liger.LigerPlugin
liger_rope: true
liger_rms_norm: true
liger_glu_activation: true
liger_layer_norm: true
liger_fused_linear_cross_entropy: true
datasets:
- path: "./train_data.jsonl"
type: chat_template
field_messages: messages
message_property_mappings:
role: role
content: content
roles_to_train: ['assistant']
train_on_eos: turn
learning_rate: 2e-5
lr_scheduler: cosine
weight_decay: 0.03
warmup_steps: 450
dataset_prepared_path:
val_set_size: 0.2
output_dir: ./outputs/out
sequence_len: 10400
sample_packing: true
pad_to_sequence_len: true
eval_sample_packing: true
wandb_project: linabot
wandb_entity:
wandb_watch: all
wandb_name:
wandb_log_model:
gradient_accumulation_steps: 1
micro_batch_size: 4
num_epochs: 5
optimizer: adalomo
lr_scheduler: cosine
learning_rate: 0.0002024
flash_attention: true
flash_attn_cross_entropy: false
flash_attn_rms_norm: true
flash_attn_fuse_qkv: false
flash_attn_fuse_mlp: true
torch_compile_mode: "max-autotune"
bf16: auto
tf32: false
gradient_checkpointing: true
resume_from_checkpoint:
logging_steps: 1
evals_per_epoch: 8
saves_per_epoch: 1
weight_decay: 0.03
special_tokens:
bos_token: "<s>"
eos_token: "</s>"
pad_token: "<pad>"
```
</details><br>
# linabot
This model is a fine-tuned version of [mistralai/Mistral-Nemo-Instruct-2407](https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407) on the ./train_data.jsonl dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4157
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0002024
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Use adalomo and the args are:
No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 450
- num_epochs: 5.0
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 0.4415 | 0.0083 | 1 | 0.5382 |
| 0.4958 | 0.125 | 15 | 0.5380 |
| 0.371 | 0.25 | 30 | 0.5371 |
| 0.4364 | 0.375 | 45 | 0.5347 |
| 0.3777 | 0.5 | 60 | 0.5309 |
| 0.3962 | 0.625 | 75 | 0.5244 |
| 0.3341 | 0.75 | 90 | 0.5168 |
| 0.3259 | 0.875 | 105 | 0.5070 |
| 0.3238 | 1.0 | 120 | 0.4966 |
| 0.36 | 1.125 | 135 | 0.4866 |
| 0.264 | 1.25 | 150 | 0.4793 |
| 0.3319 | 1.375 | 165 | 0.4714 |
| 0.3731 | 1.5 | 180 | 0.4641 |
| 0.325 | 1.625 | 195 | 0.4581 |
| 0.3477 | 1.75 | 210 | 0.4526 |
| 0.2851 | 1.875 | 225 | 0.4481 |
| 0.2732 | 2.0 | 240 | 0.4416 |
| 0.3367 | 2.125 | 255 | 0.4388 |
| 0.2605 | 2.25 | 270 | 0.4366 |
| 0.2725 | 2.375 | 285 | 0.4333 |
| 0.3374 | 2.5 | 300 | 0.4291 |
| 0.275 | 2.625 | 315 | 0.4250 |
| 0.1803 | 2.75 | 330 | 0.4214 |
| 0.3441 | 2.875 | 345 | 0.4189 |
| 0.1505 | 3.0 | 360 | 0.4172 |
| 0.2216 | 3.125 | 375 | 0.4186 |
| 0.1833 | 3.25 | 390 | 0.4185 |
| 0.2586 | 3.375 | 405 | 0.4153 |
| 0.1754 | 3.5 | 420 | 0.4152 |
| 0.2272 | 3.625 | 435 | 0.4136 |
| 0.174 | 3.75 | 450 | 0.4129 |
| 0.1794 | 3.875 | 465 | 0.4074 |
| 0.1779 | 4.0 | 480 | 0.4086 |
| 0.1752 | 4.125 | 495 | 0.4164 |
| 0.1745 | 4.25 | 510 | 0.4173 |
| 0.1314 | 4.375 | 525 | 0.4158 |
| 0.1923 | 4.5 | 540 | 0.4155 |
| 0.1848 | 4.625 | 555 | 0.4156 |
| 0.1185 | 4.75 | 570 | 0.4155 |
| 0.2255 | 4.875 | 585 | 0.4153 |
| 0.1841 | 5.0 | 600 | 0.4157 |
### Framework versions
- Transformers 4.50.3
- Pytorch 2.5.1+cu124
- Datasets 3.5.0
- Tokenizers 0.21.1
|