sedrickkeh commited on
Commit
06a90d5
·
verified ·
1 Parent(s): d9322e0

End of training

Browse files
README.md CHANGED
@@ -4,6 +4,7 @@ license: apache-2.0
4
  base_model: mistralai/Mistral-7B-v0.1
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: hp_ablations_grid_mistral_bsz4096_lr5e-6_scheduler-cosine-warmup0.15
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # hp_ablations_grid_mistral_bsz4096_lr5e-6_scheduler-cosine-warmup0.15
17
 
18
- This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: 0.0598
21
 
 
4
  base_model: mistralai/Mistral-7B-v0.1
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: hp_ablations_grid_mistral_bsz4096_lr5e-6_scheduler-cosine-warmup0.15
 
16
 
17
  # hp_ablations_grid_mistral_bsz4096_lr5e-6_scheduler-cosine-warmup0.15
18
 
19
+ This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mlfoundations-dev/oh-dcft-v3-llama3.1-nemotron-70b_shareGPT_format dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.0598
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_loss": 0.05977506935596466,
4
+ "eval_runtime": 88.9468,
5
+ "eval_samples_per_second": 203.526,
6
+ "eval_steps_per_second": 0.405,
7
+ "total_flos": 3375200049561600.0,
8
+ "train_loss": 0.5070906683566079,
9
+ "train_runtime": 15180.8601,
10
+ "train_samples_per_second": 67.971,
11
+ "train_steps_per_second": 0.017
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_loss": 0.05977506935596466,
4
+ "eval_runtime": 88.9468,
5
+ "eval_samples_per_second": 203.526,
6
+ "eval_steps_per_second": 0.405
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "total_flos": 3375200049561600.0,
4
+ "train_loss": 0.5070906683566079,
5
+ "train_runtime": 15180.8601,
6
+ "train_samples_per_second": 67.971,
7
+ "train_steps_per_second": 0.017
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,241 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 252,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.11904761904761904,
13
+ "grad_norm": 12.131371063020744,
14
+ "learning_rate": 1.3157894736842106e-06,
15
+ "loss": 0.8467,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.23809523809523808,
20
+ "grad_norm": 2.8702347939026684,
21
+ "learning_rate": 2.631578947368421e-06,
22
+ "loss": 0.7188,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.35714285714285715,
27
+ "grad_norm": 2.5230518976542506,
28
+ "learning_rate": 3.947368421052632e-06,
29
+ "loss": 0.635,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.47619047619047616,
34
+ "grad_norm": 1.6425584319246187,
35
+ "learning_rate": 4.998922515567496e-06,
36
+ "loss": 0.5914,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.5952380952380952,
41
+ "grad_norm": 3.9561567299131952,
42
+ "learning_rate": 4.9613079925074435e-06,
43
+ "loss": 0.5662,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.7142857142857143,
48
+ "grad_norm": 2.8506571881322786,
49
+ "learning_rate": 4.870744433740688e-06,
50
+ "loss": 0.5495,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.8333333333333334,
55
+ "grad_norm": 2.9284294355648997,
56
+ "learning_rate": 4.729180093253106e-06,
57
+ "loss": 0.5354,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.9523809523809523,
62
+ "grad_norm": 1.608687573355936,
63
+ "learning_rate": 4.53966038240406e-06,
64
+ "loss": 0.523,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 1.0,
69
+ "eval_loss": 0.06461798399686813,
70
+ "eval_runtime": 89.5731,
71
+ "eval_samples_per_second": 202.103,
72
+ "eval_steps_per_second": 0.402,
73
+ "step": 84
74
+ },
75
+ {
76
+ "epoch": 1.0714285714285714,
77
+ "grad_norm": 3.5277524106115705,
78
+ "learning_rate": 4.306262355332006e-06,
79
+ "loss": 0.5029,
80
+ "step": 90
81
+ },
82
+ {
83
+ "epoch": 1.1904761904761905,
84
+ "grad_norm": 2.9157076219013502,
85
+ "learning_rate": 4.034007001082985e-06,
86
+ "loss": 0.4873,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 1.3095238095238095,
91
+ "grad_norm": 1.9515685257710755,
92
+ "learning_rate": 3.7287512292828364e-06,
93
+ "loss": 0.4835,
94
+ "step": 110
95
+ },
96
+ {
97
+ "epoch": 1.4285714285714286,
98
+ "grad_norm": 3.0996510453406376,
99
+ "learning_rate": 3.39706187301784e-06,
100
+ "loss": 0.4795,
101
+ "step": 120
102
+ },
103
+ {
104
+ "epoch": 1.5476190476190477,
105
+ "grad_norm": 1.6487179646275694,
106
+ "learning_rate": 3.0460744194443658e-06,
107
+ "loss": 0.4738,
108
+ "step": 130
109
+ },
110
+ {
111
+ "epoch": 1.6666666666666665,
112
+ "grad_norm": 1.6412743986342637,
113
+ "learning_rate": 2.68333950719376e-06,
114
+ "loss": 0.4706,
115
+ "step": 140
116
+ },
117
+ {
118
+ "epoch": 1.7857142857142856,
119
+ "grad_norm": 1.756612697830082,
120
+ "learning_rate": 2.3166604928062407e-06,
121
+ "loss": 0.4675,
122
+ "step": 150
123
+ },
124
+ {
125
+ "epoch": 1.9047619047619047,
126
+ "grad_norm": 1.0971644453746074,
127
+ "learning_rate": 1.9539255805556346e-06,
128
+ "loss": 0.4633,
129
+ "step": 160
130
+ },
131
+ {
132
+ "epoch": 2.0,
133
+ "eval_loss": 0.06034243851900101,
134
+ "eval_runtime": 89.3139,
135
+ "eval_samples_per_second": 202.69,
136
+ "eval_steps_per_second": 0.403,
137
+ "step": 168
138
+ },
139
+ {
140
+ "epoch": 2.0238095238095237,
141
+ "grad_norm": 1.174978789267966,
142
+ "learning_rate": 1.6029381269821607e-06,
143
+ "loss": 0.4564,
144
+ "step": 170
145
+ },
146
+ {
147
+ "epoch": 2.142857142857143,
148
+ "grad_norm": 0.8542389334158027,
149
+ "learning_rate": 1.2712487707171645e-06,
150
+ "loss": 0.4356,
151
+ "step": 180
152
+ },
153
+ {
154
+ "epoch": 2.261904761904762,
155
+ "grad_norm": 0.7154367298759456,
156
+ "learning_rate": 9.659929989170156e-07,
157
+ "loss": 0.4335,
158
+ "step": 190
159
+ },
160
+ {
161
+ "epoch": 2.380952380952381,
162
+ "grad_norm": 0.5531641628675295,
163
+ "learning_rate": 6.93737644667995e-07,
164
+ "loss": 0.43,
165
+ "step": 200
166
+ },
167
+ {
168
+ "epoch": 2.5,
169
+ "grad_norm": 0.505015673695498,
170
+ "learning_rate": 4.6033961759594045e-07,
171
+ "loss": 0.429,
172
+ "step": 210
173
+ },
174
+ {
175
+ "epoch": 2.619047619047619,
176
+ "grad_norm": 0.48866524545920653,
177
+ "learning_rate": 2.708199067468939e-07,
178
+ "loss": 0.4311,
179
+ "step": 220
180
+ },
181
+ {
182
+ "epoch": 2.738095238095238,
183
+ "grad_norm": 0.46556587701838564,
184
+ "learning_rate": 1.2925556625931174e-07,
185
+ "loss": 0.4303,
186
+ "step": 230
187
+ },
188
+ {
189
+ "epoch": 2.857142857142857,
190
+ "grad_norm": 0.45472603633578945,
191
+ "learning_rate": 3.869200749255703e-08,
192
+ "loss": 0.4296,
193
+ "step": 240
194
+ },
195
+ {
196
+ "epoch": 2.9761904761904763,
197
+ "grad_norm": 0.43987009685051254,
198
+ "learning_rate": 1.0774844325039946e-09,
199
+ "loss": 0.4277,
200
+ "step": 250
201
+ },
202
+ {
203
+ "epoch": 3.0,
204
+ "eval_loss": 0.05977506935596466,
205
+ "eval_runtime": 88.5977,
206
+ "eval_samples_per_second": 204.328,
207
+ "eval_steps_per_second": 0.406,
208
+ "step": 252
209
+ },
210
+ {
211
+ "epoch": 3.0,
212
+ "step": 252,
213
+ "total_flos": 3375200049561600.0,
214
+ "train_loss": 0.5070906683566079,
215
+ "train_runtime": 15180.8601,
216
+ "train_samples_per_second": 67.971,
217
+ "train_steps_per_second": 0.017
218
+ }
219
+ ],
220
+ "logging_steps": 10,
221
+ "max_steps": 252,
222
+ "num_input_tokens_seen": 0,
223
+ "num_train_epochs": 3,
224
+ "save_steps": 500,
225
+ "stateful_callbacks": {
226
+ "TrainerControl": {
227
+ "args": {
228
+ "should_epoch_stop": false,
229
+ "should_evaluate": false,
230
+ "should_log": false,
231
+ "should_save": true,
232
+ "should_training_stop": true
233
+ },
234
+ "attributes": {}
235
+ }
236
+ },
237
+ "total_flos": 3375200049561600.0,
238
+ "train_batch_size": 8,
239
+ "trial_name": null,
240
+ "trial_params": null
241
+ }
training_eval_loss.png ADDED
training_loss.png ADDED