Hugging Face's logo

kazuyamaa
/
DeepSeek-R1-Distill-Qwen-14B-axolotl-int-v1.0-merged

PEFT
Safetensors
qwen2
axolotl
Generated from Trainer
Model card Files
xet
Community
DeepSeek-R1-Distill-Qwen-14B-axolotl-int-v1.0-merged
File size: 7,746 Bytes 
698668d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
 
---
library_name: peft
license: mit
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
tags:
- axolotl
- generated_from_trainer
datasets:
- kanhatakeyama/ramdom-to-fixed-multiturn-Calm3
- Aratako/Magpie-Tanuki-Qwen2.5-72B-Answered
- Aratako/magpie-qwen2.5-32b-reasoning-100k-formatted
- Aratako/magpie-reasoning-llama-nemotron-70b-100k-filtered
- Aratako/Open-Platypus-Japanese-masked-formatted
- kanhatakeyama/wizardlm8x22b-logical-math-coding-sft_additional-ja
- Aratako/magpie-ultra-v0.1-formatted
- Aratako/orca-agentinstruct-1M-v1-selected
- Aratako/Synthetic-JP-EN-Coding-Dataset-801k-50k
model-index:
- name: DeepSeek-R1-Distill-Qwen-14B-axolotl-int-v1.0
  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.dev0`
```yaml
# 学習のベースモデルに関する設定
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

# 学習後のモデルのHFへのアップロードに関する設定
hub_model_id: kazuyamaa/DeepSeek-R1-Distill-Qwen-14B-axolotl-int-v1.0
hub_strategy: "end"
push_dataset_to_hub:
hf_use_auth_token: true

# Liger Kernelの設定（学習の軽量・高速化）
plugins:
  - axolotl.integrations.liger.LigerPlugin
liger_cross_entropy: false
liger_rope: true
liger_rms_norm: true
liger_swiglu: true
liger_fused_linear_cross_entropy: true

# 量子化に関する設定
load_in_8bit: false
load_in_4bit: true

# SFTに利用するchat templateの設定
chat_template: gemma

# 学習データセットの前処理に関する設定
datasets:
  - path: kanhatakeyama/ramdom-to-fixed-multiturn-Calm3
    split: 20240806filtered[0:10000]
    type: chat_template
    field_messages: messages
    message_field_role: role
    message_field_content: content
  - path: Aratako/Magpie-Tanuki-Qwen2.5-72B-Answered
    split: train[0:10000]
    type: chat_template
    field_messages: messages
    message_field_role: role
    message_field_content: content
  - path: Aratako/magpie-qwen2.5-32b-reasoning-100k-formatted
    split: train[0:10000]
    type: chat_template
    field_messages: conversations
    message_field_role: role
    message_field_content: content
  - path: Aratako/magpie-reasoning-llama-nemotron-70b-100k-filtered
    split: train[0:10000]
    type: chat_template
    field_messages: conversations
    message_field_role: role
    message_field_content: content
  - path: Aratako/Open-Platypus-Japanese-masked-formatted
    split: train[0:10000]
    type: chat_template
    field_messages: conversations
    message_field_role: role
    message_field_content: content
  - path: kanhatakeyama/wizardlm8x22b-logical-math-coding-sft_additional-ja
    split: train[0:10000]
    type: chat_template
    field_messages: messages
    message_field_role: role
    message_field_content: content
  - path: Aratako/magpie-ultra-v0.1-formatted
    split: train[0:10000]
    type: chat_template
    field_messages: conversations
    message_field_role: role
    message_field_content: content
  - path: Aratako/orca-agentinstruct-1M-v1-selected
    split: train[0:10000]
    type: chat_template
    field_messages: messages
    message_field_role: role
    message_field_content: content
  - path: Aratako/Synthetic-JP-EN-Coding-Dataset-801k-50k
    split: train[0:10000]
    type: chat_template
    field_messages: messages
    message_field_role: role
    message_field_content: content

# データセット、モデルの出力先に関する設定
shuffle_merged_datasets: true
dataset_prepared_path: /workspace/data/sft-data
output_dir: /workspace/data/models/DeepSeek-R1-Distill-Qwen-14B-axolotl-int-v1.0

# valid datasetのサイズ
val_set_size: 0.05

# LoRAに関する設定（フルファインチューニングしたい場合は全て空欄にする）
adapter: qlora
lora_model_dir:
lora_r: 16
lora_alpha: 32
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:

# wandbに関する設定
wandb_project: axolotl
wandb_entity: kazukitakayamas051-securities-companies
wandb_watch:
wandb_name: sft-lora-1
wandb_log_model:

# 学習に関する様々な設定
sequence_len: 4096
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true

gradient_accumulation_steps: 16
micro_batch_size: 1
num_epochs: 1
optimizer: paged_adamw_8bit
lr_scheduler: cosine
cosine_min_lr_ratio: 0.1
learning_rate: 3e-4

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false


gradient_checkpointing: false
early_stopping_patience:
auto_resume_from_checkpoints: true
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

save_strategy: steps
save_steps: 50
save_total_limit: 2

warmup_steps: 10
eval_steps: 50
eval_batch_size: 1
eval_table_size:
eval_max_new_tokens:
debug:
deepspeed: /workspace/axolotl/deepspeed_configs/zero3_bf16.json
weight_decay: 0.01
fsdp:
fsdp_config:
special_tokens:
  pad_token: <pad>
```

</details><br>

# DeepSeek-R1-Distill-Qwen-14B-axolotl-int-v1.0

This model is a fine-tuned version of [deepseek-ai/DeepSeek-R1-Distill-Qwen-14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B) on the kanhatakeyama/ramdom-to-fixed-multiturn-Calm3, the Aratako/Magpie-Tanuki-Qwen2.5-72B-Answered, the Aratako/magpie-qwen2.5-32b-reasoning-100k-formatted, the Aratako/magpie-reasoning-llama-nemotron-70b-100k-filtered, the Aratako/Open-Platypus-Japanese-masked-formatted, the kanhatakeyama/wizardlm8x22b-logical-math-coding-sft_additional-ja, the Aratako/magpie-ultra-v0.1-formatted, the Aratako/orca-agentinstruct-1M-v1-selected and the Aratako/Synthetic-JP-EN-Coding-Dataset-801k-50k datasets.
It achieves the following results on the evaluation set:
- Loss: 0.6711

## 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: 0.0003
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- total_eval_batch_size: 2
- optimizer: Use paged_adamw_8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 1.0

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 1.1079        | 0.0015 | 1    | 1.0631          |
| 0.8387        | 0.0763 | 50   | 0.7640          |
| 0.7109        | 0.1526 | 100  | 0.7312          |
| 0.7324        | 0.2289 | 150  | 0.7155          |
| 0.8239        | 0.3051 | 200  | 0.7045          |
| 0.7019        | 0.3814 | 250  | 0.6967          |
| 0.8834        | 0.4577 | 300  | 0.6910          |
| 0.7097        | 0.5340 | 350  | 0.6857          |
| 0.6659        | 0.6103 | 400  | 0.6821          |
| 0.6755        | 0.6866 | 450  | 0.6785          |
| 0.6465        | 0.7628 | 500  | 0.6755          |
| 0.6697        | 0.8391 | 550  | 0.6735          |
| 0.8425        | 0.9154 | 600  | 0.6720          |
| 0.6461        | 0.9917 | 650  | 0.6711          |


### Framework versions

- PEFT 0.14.0
- Transformers 4.49.0
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.1