Commit
•
9044983
1
Parent(s):
59d3e97
End of training
Browse files- README.md +3 -2
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +366 -0
README.md
CHANGED
@@ -2,6 +2,7 @@
|
|
2 |
library_name: peft
|
3 |
tags:
|
4 |
- llama-factory
|
|
|
5 |
- generated_from_trainer
|
6 |
base_model: meta-llama/Llama-2-7b-hf
|
7 |
model-index:
|
@@ -14,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
14 |
|
15 |
# Llama2_AAID_new_mixed_train_final
|
16 |
|
17 |
-
This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on
|
18 |
It achieves the following results on the evaluation set:
|
19 |
-
- Loss: 0.
|
20 |
|
21 |
## Model description
|
22 |
|
|
|
2 |
library_name: peft
|
3 |
tags:
|
4 |
- llama-factory
|
5 |
+
- lora
|
6 |
- generated_from_trainer
|
7 |
base_model: meta-llama/Llama-2-7b-hf
|
8 |
model-index:
|
|
|
15 |
|
16 |
# Llama2_AAID_new_mixed_train_final
|
17 |
|
18 |
+
This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the AAID_new_mixed dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 0.5070
|
21 |
|
22 |
## Model description
|
23 |
|
all_results.json
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.22969647251845776,
|
3 |
+
"eval_loss": 0.5069935321807861,
|
4 |
+
"eval_runtime": 234.3621,
|
5 |
+
"eval_samples_per_second": 20.106,
|
6 |
+
"eval_steps_per_second": 0.632,
|
7 |
+
"total_flos": 3.9689172647569e+17,
|
8 |
+
"train_loss": 0.36669377031780426,
|
9 |
+
"train_runtime": 15000.4627,
|
10 |
+
"train_samples_per_second": 46.807,
|
11 |
+
"train_steps_per_second": 0.183
|
12 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.22969647251845776,
|
3 |
+
"eval_loss": 0.5069935321807861,
|
4 |
+
"eval_runtime": 234.3621,
|
5 |
+
"eval_samples_per_second": 20.106,
|
6 |
+
"eval_steps_per_second": 0.632
|
7 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.22969647251845776,
|
3 |
+
"total_flos": 3.9689172647569e+17,
|
4 |
+
"train_loss": 0.36669377031780426,
|
5 |
+
"train_runtime": 15000.4627,
|
6 |
+
"train_samples_per_second": 46.807,
|
7 |
+
"train_steps_per_second": 0.183
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,366 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.5069935321807861,
|
3 |
+
"best_model_checkpoint": "saves/Llama2-7B/lora/train_1/checkpoint-110",
|
4 |
+
"epoch": 0.22969647251845776,
|
5 |
+
"eval_steps": 10,
|
6 |
+
"global_step": 210,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.010937927262783703,
|
13 |
+
"grad_norm": 0.5937469005584717,
|
14 |
+
"learning_rate": 0.00029999015487222375,
|
15 |
+
"loss": 1.2211,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.010937927262783703,
|
20 |
+
"eval_loss": 0.550807535648346,
|
21 |
+
"eval_runtime": 234.7461,
|
22 |
+
"eval_samples_per_second": 20.073,
|
23 |
+
"eval_steps_per_second": 0.63,
|
24 |
+
"step": 10
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.021875854525567406,
|
28 |
+
"grad_norm": 0.4648227393627167,
|
29 |
+
"learning_rate": 0.00029996062078124905,
|
30 |
+
"loss": 0.446,
|
31 |
+
"step": 20
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.021875854525567406,
|
35 |
+
"eval_loss": 0.5179979205131531,
|
36 |
+
"eval_runtime": 234.6313,
|
37 |
+
"eval_samples_per_second": 20.083,
|
38 |
+
"eval_steps_per_second": 0.631,
|
39 |
+
"step": 20
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.03281378178835111,
|
43 |
+
"grad_norm": 0.42532142996788025,
|
44 |
+
"learning_rate": 0.0002999114016039678,
|
45 |
+
"loss": 0.3889,
|
46 |
+
"step": 30
|
47 |
+
},
|
48 |
+
{
|
49 |
+
"epoch": 0.03281378178835111,
|
50 |
+
"eval_loss": 0.5139310956001282,
|
51 |
+
"eval_runtime": 234.5844,
|
52 |
+
"eval_samples_per_second": 20.087,
|
53 |
+
"eval_steps_per_second": 0.631,
|
54 |
+
"step": 30
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 0.04375170905113481,
|
58 |
+
"grad_norm": 0.47056564688682556,
|
59 |
+
"learning_rate": 0.00029984250380130117,
|
60 |
+
"loss": 0.3757,
|
61 |
+
"step": 40
|
62 |
+
},
|
63 |
+
{
|
64 |
+
"epoch": 0.04375170905113481,
|
65 |
+
"eval_loss": 0.575846791267395,
|
66 |
+
"eval_runtime": 234.7308,
|
67 |
+
"eval_samples_per_second": 20.074,
|
68 |
+
"eval_steps_per_second": 0.631,
|
69 |
+
"step": 40
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.05468963631391851,
|
73 |
+
"grad_norm": 0.2359870821237564,
|
74 |
+
"learning_rate": 0.0002997539364173515,
|
75 |
+
"loss": 0.3551,
|
76 |
+
"step": 50
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 0.05468963631391851,
|
80 |
+
"eval_loss": 0.5461050271987915,
|
81 |
+
"eval_runtime": 234.7378,
|
82 |
+
"eval_samples_per_second": 20.073,
|
83 |
+
"eval_steps_per_second": 0.63,
|
84 |
+
"step": 50
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 0.06562756357670221,
|
88 |
+
"grad_norm": 0.19186203181743622,
|
89 |
+
"learning_rate": 0.00029964571107821494,
|
90 |
+
"loss": 0.3351,
|
91 |
+
"step": 60
|
92 |
+
},
|
93 |
+
{
|
94 |
+
"epoch": 0.06562756357670221,
|
95 |
+
"eval_loss": 0.5407155752182007,
|
96 |
+
"eval_runtime": 234.7344,
|
97 |
+
"eval_samples_per_second": 20.074,
|
98 |
+
"eval_steps_per_second": 0.63,
|
99 |
+
"step": 60
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.07656549083948591,
|
103 |
+
"grad_norm": 0.25036656856536865,
|
104 |
+
"learning_rate": 0.00029951784199045534,
|
105 |
+
"loss": 0.3269,
|
106 |
+
"step": 70
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 0.07656549083948591,
|
110 |
+
"eval_loss": 0.5335067510604858,
|
111 |
+
"eval_runtime": 234.559,
|
112 |
+
"eval_samples_per_second": 20.089,
|
113 |
+
"eval_steps_per_second": 0.631,
|
114 |
+
"step": 70
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.08750341810226962,
|
118 |
+
"grad_norm": 0.3150342106819153,
|
119 |
+
"learning_rate": 0.0002993703459392396,
|
120 |
+
"loss": 0.3231,
|
121 |
+
"step": 80
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.08750341810226962,
|
125 |
+
"eval_loss": 0.5120783448219299,
|
126 |
+
"eval_runtime": 234.641,
|
127 |
+
"eval_samples_per_second": 20.082,
|
128 |
+
"eval_steps_per_second": 0.631,
|
129 |
+
"step": 80
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.09844134536505332,
|
133 |
+
"grad_norm": 0.3419171869754791,
|
134 |
+
"learning_rate": 0.00029920324228613376,
|
135 |
+
"loss": 0.3324,
|
136 |
+
"step": 90
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.09844134536505332,
|
140 |
+
"eval_loss": 0.5196456909179688,
|
141 |
+
"eval_runtime": 234.6621,
|
142 |
+
"eval_samples_per_second": 20.08,
|
143 |
+
"eval_steps_per_second": 0.631,
|
144 |
+
"step": 90
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 0.10937927262783702,
|
148 |
+
"grad_norm": 0.2197423279285431,
|
149 |
+
"learning_rate": 0.0002990165529665622,
|
150 |
+
"loss": 0.3039,
|
151 |
+
"step": 100
|
152 |
+
},
|
153 |
+
{
|
154 |
+
"epoch": 0.10937927262783702,
|
155 |
+
"eval_loss": 0.5251961350440979,
|
156 |
+
"eval_runtime": 234.8661,
|
157 |
+
"eval_samples_per_second": 20.062,
|
158 |
+
"eval_steps_per_second": 0.63,
|
159 |
+
"step": 100
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 0.12031719989062073,
|
163 |
+
"grad_norm": 0.16334187984466553,
|
164 |
+
"learning_rate": 0.0002988103024869277,
|
165 |
+
"loss": 0.3083,
|
166 |
+
"step": 110
|
167 |
+
},
|
168 |
+
{
|
169 |
+
"epoch": 0.12031719989062073,
|
170 |
+
"eval_loss": 0.5069935321807861,
|
171 |
+
"eval_runtime": 234.5526,
|
172 |
+
"eval_samples_per_second": 20.089,
|
173 |
+
"eval_steps_per_second": 0.631,
|
174 |
+
"step": 110
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 0.13125512715340443,
|
178 |
+
"grad_norm": 0.2504599392414093,
|
179 |
+
"learning_rate": 0.00029858451792139453,
|
180 |
+
"loss": 0.3034,
|
181 |
+
"step": 120
|
182 |
+
},
|
183 |
+
{
|
184 |
+
"epoch": 0.13125512715340443,
|
185 |
+
"eval_loss": 0.5375232100486755,
|
186 |
+
"eval_runtime": 234.9066,
|
187 |
+
"eval_samples_per_second": 20.059,
|
188 |
+
"eval_steps_per_second": 0.63,
|
189 |
+
"step": 120
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 0.14219305441618812,
|
193 |
+
"grad_norm": 0.20455080270767212,
|
194 |
+
"learning_rate": 0.0002983392289083346,
|
195 |
+
"loss": 0.3061,
|
196 |
+
"step": 130
|
197 |
+
},
|
198 |
+
{
|
199 |
+
"epoch": 0.14219305441618812,
|
200 |
+
"eval_loss": 0.5345537066459656,
|
201 |
+
"eval_runtime": 234.955,
|
202 |
+
"eval_samples_per_second": 20.055,
|
203 |
+
"eval_steps_per_second": 0.63,
|
204 |
+
"step": 130
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 0.15313098167897182,
|
208 |
+
"grad_norm": 0.20371787250041962,
|
209 |
+
"learning_rate": 0.0002980744676464371,
|
210 |
+
"loss": 0.3021,
|
211 |
+
"step": 140
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"epoch": 0.15313098167897182,
|
215 |
+
"eval_loss": 0.5111611485481262,
|
216 |
+
"eval_runtime": 234.7152,
|
217 |
+
"eval_samples_per_second": 20.075,
|
218 |
+
"eval_steps_per_second": 0.631,
|
219 |
+
"step": 140
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.16406890894175555,
|
223 |
+
"grad_norm": 0.14063598215579987,
|
224 |
+
"learning_rate": 0.0002977902688904813,
|
225 |
+
"loss": 0.3067,
|
226 |
+
"step": 150
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.16406890894175555,
|
230 |
+
"eval_loss": 0.5191352367401123,
|
231 |
+
"eval_runtime": 234.5955,
|
232 |
+
"eval_samples_per_second": 20.086,
|
233 |
+
"eval_steps_per_second": 0.631,
|
234 |
+
"step": 150
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.17500683620453925,
|
238 |
+
"grad_norm": 0.20113405585289001,
|
239 |
+
"learning_rate": 0.00029748666994677467,
|
240 |
+
"loss": 0.2958,
|
241 |
+
"step": 160
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.17500683620453925,
|
245 |
+
"eval_loss": 0.5278245210647583,
|
246 |
+
"eval_runtime": 234.5747,
|
247 |
+
"eval_samples_per_second": 20.087,
|
248 |
+
"eval_steps_per_second": 0.631,
|
249 |
+
"step": 160
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 0.18594476346732294,
|
253 |
+
"grad_norm": 0.20178332924842834,
|
254 |
+
"learning_rate": 0.00029716371066825593,
|
255 |
+
"loss": 0.3002,
|
256 |
+
"step": 170
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"epoch": 0.18594476346732294,
|
260 |
+
"eval_loss": 0.5170465111732483,
|
261 |
+
"eval_runtime": 234.6689,
|
262 |
+
"eval_samples_per_second": 20.079,
|
263 |
+
"eval_steps_per_second": 0.631,
|
264 |
+
"step": 170
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 0.19688269073010664,
|
268 |
+
"grad_norm": 0.19617249071598053,
|
269 |
+
"learning_rate": 0.0002968214334492632,
|
270 |
+
"loss": 0.2896,
|
271 |
+
"step": 180
|
272 |
+
},
|
273 |
+
{
|
274 |
+
"epoch": 0.19688269073010664,
|
275 |
+
"eval_loss": 0.5085877180099487,
|
276 |
+
"eval_runtime": 234.6318,
|
277 |
+
"eval_samples_per_second": 20.083,
|
278 |
+
"eval_steps_per_second": 0.631,
|
279 |
+
"step": 180
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 0.20782061799289034,
|
283 |
+
"grad_norm": 0.2610602080821991,
|
284 |
+
"learning_rate": 0.00029645988321996917,
|
285 |
+
"loss": 0.2989,
|
286 |
+
"step": 190
|
287 |
+
},
|
288 |
+
{
|
289 |
+
"epoch": 0.20782061799289034,
|
290 |
+
"eval_loss": 0.530036985874176,
|
291 |
+
"eval_runtime": 234.599,
|
292 |
+
"eval_samples_per_second": 20.085,
|
293 |
+
"eval_steps_per_second": 0.631,
|
294 |
+
"step": 190
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"epoch": 0.21875854525567404,
|
298 |
+
"grad_norm": 0.1975240856409073,
|
299 |
+
"learning_rate": 0.00029607910744048336,
|
300 |
+
"loss": 0.3012,
|
301 |
+
"step": 200
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"epoch": 0.21875854525567404,
|
305 |
+
"eval_loss": 0.5137269496917725,
|
306 |
+
"eval_runtime": 234.4763,
|
307 |
+
"eval_samples_per_second": 20.096,
|
308 |
+
"eval_steps_per_second": 0.631,
|
309 |
+
"step": 200
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 0.22969647251845776,
|
313 |
+
"grad_norm": 0.19913341104984283,
|
314 |
+
"learning_rate": 0.00029567915609462174,
|
315 |
+
"loss": 0.2802,
|
316 |
+
"step": 210
|
317 |
+
},
|
318 |
+
{
|
319 |
+
"epoch": 0.22969647251845776,
|
320 |
+
"eval_loss": 0.5159074068069458,
|
321 |
+
"eval_runtime": 234.549,
|
322 |
+
"eval_samples_per_second": 20.09,
|
323 |
+
"eval_steps_per_second": 0.631,
|
324 |
+
"step": 210
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.22969647251845776,
|
328 |
+
"step": 210,
|
329 |
+
"total_flos": 3.9689172647569e+17,
|
330 |
+
"train_loss": 0.36669377031780426,
|
331 |
+
"train_runtime": 15000.4627,
|
332 |
+
"train_samples_per_second": 46.807,
|
333 |
+
"train_steps_per_second": 0.183
|
334 |
+
}
|
335 |
+
],
|
336 |
+
"logging_steps": 10,
|
337 |
+
"max_steps": 2742,
|
338 |
+
"num_input_tokens_seen": 0,
|
339 |
+
"num_train_epochs": 3,
|
340 |
+
"save_steps": 10,
|
341 |
+
"stateful_callbacks": {
|
342 |
+
"EarlyStoppingCallback": {
|
343 |
+
"args": {
|
344 |
+
"early_stopping_patience": 10,
|
345 |
+
"early_stopping_threshold": 0.0
|
346 |
+
},
|
347 |
+
"attributes": {
|
348 |
+
"early_stopping_patience_counter": 0
|
349 |
+
}
|
350 |
+
},
|
351 |
+
"TrainerControl": {
|
352 |
+
"args": {
|
353 |
+
"should_epoch_stop": false,
|
354 |
+
"should_evaluate": false,
|
355 |
+
"should_log": false,
|
356 |
+
"should_save": true,
|
357 |
+
"should_training_stop": true
|
358 |
+
},
|
359 |
+
"attributes": {}
|
360 |
+
}
|
361 |
+
},
|
362 |
+
"total_flos": 3.9689172647569e+17,
|
363 |
+
"train_batch_size": 32,
|
364 |
+
"trial_name": null,
|
365 |
+
"trial_params": null
|
366 |
+
}
|