Hugging Face's logo

neginashz
/
qlora-qwen-25-7b-instruct-s

PEFT
Safetensors
qwen2
axolotl
Generated from Trainer
4-bit precision
bitsandbytes
Model card Files Community
neginashz commited on
Commit
98e2641
·
verified ·
1 Parent(s): a704933

Model save

Browse files
Files changed (1) hide show
  1. README.md +176 -0
README.md ADDED
@@ -0,0 +1,176 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: apache-2.0
4
+ base_model: Qwen/Qwen2.5-7B-Instruct
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ datasets:
9
+ - medalpaca/medical_meadow_medqa
10
+ model-index:
11
+ - name: qlora-qwen-25-7b-instruct-s
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ [<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)
19
+ <details><summary>See axolotl config</summary>
20
+
21
+ axolotl version: `0.6.0`
22
+ ```yaml
23
+ base_model: Qwen/Qwen2.5-7B-Instruct
24
+ trust_remote_code: true
25
+
26
+ load_in_8bit: false
27
+ load_in_4bit: true
28
+ strict: false
29
+
30
+ datasets:
31
+ - path: medalpaca/medical_meadow_medqa
32
+ type: alpaca
33
+ dataset_prepared_path:
34
+ val_set_size: 0.1
35
+ output_dir: ./qlora-qwen25-instruct
36
+
37
+ sequence_len: 2048
38
+ sample_packing: true
39
+ eval_sample_packing: true
40
+ pad_to_sequence_len: true
41
+
42
+ adapter: qlora
43
+ lora_model_dir:
44
+ lora_r: 256
45
+ lora_alpha: 256
46
+ lora_dropout: 0.05
47
+ lora_target_linear: true
48
+ lora_fan_in_fan_out:
49
+
50
+ wandb_project:
51
+ wandb_entity:
52
+ wandb_watch:
53
+ wandb_name:
54
+ wandb_log_model:
55
+
56
+ gradient_accumulation_steps: 1
57
+ micro_batch_size: 2
58
+ num_epochs: 3
59
+ optimizer: adamw_torch
60
+ lr_scheduler: cosine
61
+ learning_rate: 0.00002
62
+
63
+ train_on_inputs: false
64
+ group_by_length: false
65
+ bf16: true
66
+ fp16:
67
+ tf32:
68
+
69
+ gradient_checkpointing: true
70
+ gradient_checkpointing_kwargs:
71
+ use_reentrant: false
72
+ early_stopping_patience:
73
+ resume_from_checkpoint:
74
+ local_rank:
75
+ logging_steps: 1
76
+ xformers_attention:
77
+ flash_attention: true
78
+
79
+ warmup_steps:
80
+ evals_per_epoch: 4
81
+ saves_per_epoch: 1
82
+ debug:
83
+ deepspeed:
84
+ weight_decay: 0.0
85
+ fsdp:
86
+ - full_shard
87
+ - auto_wrap
88
+ fsdp_config:
89
+ fsdp_limit_all_gathers: true
90
+ fsdp_sync_module_states: true
91
+ fsdp_offload_params: true
92
+ fsdp_use_orig_params: false
93
+ fsdp_cpu_ram_efficient_loading: true
94
+ fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
95
+ fsdp_transformer_layer_cls_to_wrap: Qwen2DecoderLayer
96
+ fsdp_state_dict_type: FULL_STATE_DICT
97
+ fsdp_sharding_strategy: FULL_SHARD
98
+ special_tokens:
99
+
100
+ wandb_project: qlora-qwen-25-7b-instruct
101
+ wandb_entity:
102
+ wandb_watch:
103
+ wandb_name:
104
+ wandb_log_model:
105
+
106
+ hub_model_id: neginashz/qlora-qwen-25-7b-instruct-s
107
+ hub_strategy:
108
+ early_stopping_patience:
109
+
110
+ resume_from_checkpoint:
111
+ auto_resume_from_checkpoints: true
112
+
113
+ ```
114
+
115
+ </details><br>
116
+
117
+ # qlora-qwen-25-7b-instruct-s
118
+
119
+ This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the medalpaca/medical_meadow_medqa dataset.
120
+ It achieves the following results on the evaluation set:
121
+ - Loss: 0.1608
122
+
123
+ ## Model description
124
+
125
+ More information needed
126
+
127
+ ## Intended uses & limitations
128
+
129
+ More information needed
130
+
131
+ ## Training and evaluation data
132
+
133
+ More information needed
134
+
135
+ ## Training procedure
136
+
137
+ ### Training hyperparameters
138
+
139
+ The following hyperparameters were used during training:
140
+ - learning_rate: 2e-05
141
+ - train_batch_size: 2
142
+ - eval_batch_size: 2
143
+ - seed: 42
144
+ - distributed_type: multi-GPU
145
+ - num_devices: 4
146
+ - total_train_batch_size: 8
147
+ - total_eval_batch_size: 8
148
+ - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
149
+ - lr_scheduler_type: cosine
150
+ - lr_scheduler_warmup_steps: 14
151
+ - num_epochs: 3
152
+
153
+ ### Training results
154
+
155
+ | Training Loss | Epoch | Step | Validation Loss |
156
+ |:-------------:|:------:|:----:|:---------------:|
157
+ | 0.1216 | 0.2530 | 42 | 0.1267 |
158
+ | 0.1366 | 0.5060 | 84 | 0.1142 |
159
+ | 0.0914 | 0.7590 | 126 | 0.1104 |
160
+ | 0.0814 | 1.0060 | 168 | 0.1050 |
161
+ | 0.0763 | 1.2590 | 210 | 0.1113 |
162
+ | 0.0746 | 1.5120 | 252 | 0.1147 |
163
+ | 0.0467 | 1.7651 | 294 | 0.1125 |
164
+ | 0.0176 | 2.0120 | 336 | 0.1154 |
165
+ | 0.0367 | 2.2651 | 378 | 0.1605 |
166
+ | 0.0349 | 2.5181 | 420 | 0.1571 |
167
+ | 0.0173 | 2.7711 | 462 | 0.1608 |
168
+
169
+
170
+ ### Framework versions
171
+
172
+ - PEFT 0.14.0
173
+ - Transformers 4.47.0
174
+ - Pytorch 2.5.1+cu124
175
+ - Datasets 3.1.0
176
+ - Tokenizers 0.21.0