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