MHGanainy/gpt2-xl-lora-multi-5
Browse files- README.md +2 -0
- all_results.json +10 -10
- eval_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +322 -161
README.md
CHANGED
@@ -15,6 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
# gpt2-xl-lora-multi-5
|
16 |
|
17 |
This model is a fine-tuned version of [openai-community/gpt2-xl](https://huggingface.co/openai-community/gpt2-xl) on an unknown dataset.
|
|
|
|
|
18 |
|
19 |
## Model description
|
20 |
|
|
|
15 |
# gpt2-xl-lora-multi-5
|
16 |
|
17 |
This model is a fine-tuned version of [openai-community/gpt2-xl](https://huggingface.co/openai-community/gpt2-xl) on an unknown dataset.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 2.4173
|
20 |
|
21 |
## Model description
|
22 |
|
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 2.
|
4 |
-
"eval_runtime":
|
5 |
-
"eval_samples_per_second":
|
6 |
-
"eval_steps_per_second": 0.
|
7 |
-
"perplexity":
|
8 |
-
"total_flos":
|
9 |
-
"train_loss": 2.
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second":
|
12 |
-
"train_steps_per_second": 3.
|
13 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 2.4172921180725098,
|
4 |
+
"eval_runtime": 1303.5259,
|
5 |
+
"eval_samples_per_second": 8.509,
|
6 |
+
"eval_steps_per_second": 0.532,
|
7 |
+
"perplexity": 11.21544804932504,
|
8 |
+
"total_flos": 8.917940750474281e+17,
|
9 |
+
"train_loss": 2.4952426230985356,
|
10 |
+
"train_runtime": 1729.9843,
|
11 |
+
"train_samples_per_second": 56.6,
|
12 |
+
"train_steps_per_second": 3.538
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 2.
|
4 |
-
"eval_runtime":
|
5 |
-
"eval_samples_per_second":
|
6 |
-
"eval_steps_per_second": 0.
|
7 |
-
"perplexity":
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 2.4172921180725098,
|
4 |
+
"eval_runtime": 1303.5259,
|
5 |
+
"eval_samples_per_second": 8.509,
|
6 |
+
"eval_steps_per_second": 0.532,
|
7 |
+
"perplexity": 11.21544804932504
|
8 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"total_flos":
|
4 |
-
"train_loss": 2.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 3.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"total_flos": 8.917940750474281e+17,
|
4 |
+
"train_loss": 2.4952426230985356,
|
5 |
+
"train_runtime": 1729.9843,
|
6 |
+
"train_samples_per_second": 56.6,
|
7 |
+
"train_steps_per_second": 3.538
|
8 |
}
|
trainer_state.json
CHANGED
@@ -3,289 +3,450 @@
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
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.
|
13 |
-
"grad_norm": 0.
|
14 |
-
"learning_rate":
|
15 |
-
"loss":
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm": 0.
|
21 |
-
"learning_rate":
|
22 |
-
"loss":
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm": 0.
|
28 |
-
"learning_rate": 1.
|
29 |
-
"loss":
|
30 |
"step": 300
|
31 |
},
|
32 |
{
|
33 |
-
"epoch": 0.
|
34 |
-
"grad_norm": 0.
|
35 |
-
"learning_rate":
|
36 |
-
"loss":
|
37 |
"step": 400
|
38 |
},
|
39 |
{
|
40 |
-
"epoch": 0.
|
41 |
-
"grad_norm":
|
42 |
-
"learning_rate":
|
43 |
-
"loss":
|
44 |
"step": 500
|
45 |
},
|
46 |
{
|
47 |
-
"epoch": 0.
|
48 |
-
"grad_norm": 0.
|
49 |
-
"learning_rate":
|
50 |
-
"loss":
|
51 |
"step": 600
|
52 |
},
|
53 |
{
|
54 |
-
"epoch": 0.
|
55 |
-
"grad_norm": 0.
|
56 |
-
"learning_rate":
|
57 |
-
"loss": 2.
|
58 |
"step": 700
|
59 |
},
|
60 |
{
|
61 |
-
"epoch": 0.
|
62 |
-
"grad_norm": 0.
|
63 |
-
"learning_rate":
|
64 |
-
"loss": 2.
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
-
"epoch": 0.
|
69 |
-
"grad_norm": 0.
|
70 |
-
"learning_rate":
|
71 |
-
"loss": 2.
|
72 |
"step": 900
|
73 |
},
|
74 |
{
|
75 |
-
"epoch": 0.
|
76 |
-
"grad_norm": 0.
|
77 |
-
"learning_rate":
|
78 |
-
"loss": 2.
|
79 |
"step": 1000
|
80 |
},
|
81 |
{
|
82 |
-
"epoch": 0.
|
83 |
-
"grad_norm": 0.
|
84 |
-
"learning_rate":
|
85 |
-
"loss": 2.
|
86 |
"step": 1100
|
87 |
},
|
88 |
{
|
89 |
-
"epoch": 0.
|
90 |
-
"grad_norm": 0.
|
91 |
-
"learning_rate":
|
92 |
-
"loss": 2.
|
93 |
"step": 1200
|
94 |
},
|
95 |
{
|
96 |
-
"epoch": 0.
|
97 |
-
"grad_norm": 0.
|
98 |
-
"learning_rate":
|
99 |
-
"loss": 2.
|
100 |
"step": 1300
|
101 |
},
|
102 |
{
|
103 |
-
"epoch": 0.
|
104 |
-
"grad_norm": 0.
|
105 |
-
"learning_rate":
|
106 |
-
"loss": 2.
|
107 |
"step": 1400
|
108 |
},
|
109 |
{
|
110 |
-
"epoch": 0.
|
111 |
-
"grad_norm": 0.
|
112 |
-
"learning_rate":
|
113 |
-
"loss": 2.
|
114 |
"step": 1500
|
115 |
},
|
116 |
{
|
117 |
-
"epoch": 0.
|
118 |
-
"grad_norm": 0.
|
119 |
-
"learning_rate":
|
120 |
-
"loss": 2.
|
121 |
"step": 1600
|
122 |
},
|
123 |
{
|
124 |
-
"epoch": 0.
|
125 |
-
"grad_norm": 0.
|
126 |
-
"learning_rate":
|
127 |
-
"loss": 2.
|
128 |
"step": 1700
|
129 |
},
|
130 |
{
|
131 |
-
"epoch": 0.
|
132 |
-
"grad_norm": 0.
|
133 |
-
"learning_rate":
|
134 |
-
"loss": 2.
|
135 |
"step": 1800
|
136 |
},
|
137 |
{
|
138 |
-
"epoch": 0.
|
139 |
-
"grad_norm": 0.
|
140 |
-
"learning_rate":
|
141 |
-
"loss": 2.
|
142 |
"step": 1900
|
143 |
},
|
144 |
{
|
145 |
-
"epoch": 0.
|
146 |
-
"grad_norm": 0.
|
147 |
-
"learning_rate":
|
148 |
-
"loss": 2.
|
149 |
"step": 2000
|
150 |
},
|
151 |
{
|
152 |
-
"epoch": 0.
|
153 |
-
"grad_norm": 0.
|
154 |
-
"learning_rate":
|
155 |
-
"loss": 2.
|
156 |
"step": 2100
|
157 |
},
|
158 |
{
|
159 |
-
"epoch": 0.
|
160 |
-
"grad_norm": 0.
|
161 |
-
"learning_rate":
|
162 |
-
"loss": 2.
|
163 |
"step": 2200
|
164 |
},
|
165 |
{
|
166 |
-
"epoch": 0.
|
167 |
-
"grad_norm": 0.
|
168 |
-
"learning_rate":
|
169 |
-
"loss": 2.
|
170 |
"step": 2300
|
171 |
},
|
172 |
{
|
173 |
-
"epoch": 0.
|
174 |
-
"grad_norm": 0.
|
175 |
-
"learning_rate":
|
176 |
-
"loss": 2.
|
177 |
"step": 2400
|
178 |
},
|
179 |
{
|
180 |
-
"epoch": 0.
|
181 |
-
"grad_norm": 0.
|
182 |
-
"learning_rate": 1.
|
183 |
-
"loss": 2.
|
184 |
"step": 2500
|
185 |
},
|
186 |
{
|
187 |
-
"epoch": 0.
|
188 |
-
"grad_norm": 0.
|
189 |
-
"learning_rate": 1.
|
190 |
-
"loss": 2.
|
191 |
"step": 2600
|
192 |
},
|
193 |
{
|
194 |
-
"epoch": 0.
|
195 |
-
"grad_norm": 0.
|
196 |
-
"learning_rate": 1.
|
197 |
-
"loss": 2.
|
198 |
"step": 2700
|
199 |
},
|
200 |
{
|
201 |
-
"epoch": 0.
|
202 |
-
"grad_norm": 0.
|
203 |
-
"learning_rate": 1.
|
204 |
-
"loss": 2.
|
205 |
"step": 2800
|
206 |
},
|
207 |
{
|
208 |
-
"epoch": 0.
|
209 |
-
"grad_norm": 0.
|
210 |
-
"learning_rate": 1.
|
211 |
-
"loss": 2.
|
212 |
"step": 2900
|
213 |
},
|
214 |
{
|
215 |
-
"epoch": 0.
|
216 |
-
"grad_norm": 0.
|
217 |
-
"learning_rate": 1.
|
218 |
-
"loss": 2.
|
219 |
"step": 3000
|
220 |
},
|
221 |
{
|
222 |
-
"epoch": 0.
|
223 |
-
"grad_norm": 0.
|
224 |
-
"learning_rate": 1.
|
225 |
-
"loss": 2.
|
226 |
"step": 3100
|
227 |
},
|
228 |
{
|
229 |
-
"epoch": 0.
|
230 |
-
"grad_norm": 0.
|
231 |
-
"learning_rate": 1.
|
232 |
-
"loss": 2.
|
233 |
"step": 3200
|
234 |
},
|
235 |
{
|
236 |
-
"epoch": 0.
|
237 |
-
"grad_norm": 0.
|
238 |
-
"learning_rate": 1.
|
239 |
-
"loss": 2.
|
240 |
"step": 3300
|
241 |
},
|
242 |
{
|
243 |
-
"epoch": 0.
|
244 |
-
"grad_norm": 0.
|
245 |
-
"learning_rate": 1.
|
246 |
-
"loss": 2.
|
247 |
"step": 3400
|
248 |
},
|
249 |
{
|
250 |
-
"epoch": 0.
|
251 |
-
"grad_norm": 0.
|
252 |
-
"learning_rate":
|
253 |
-
"loss": 2.
|
254 |
"step": 3500
|
255 |
},
|
256 |
{
|
257 |
-
"epoch": 0.
|
258 |
-
"grad_norm": 0.
|
259 |
-
"learning_rate":
|
260 |
-
"loss": 2.
|
261 |
"step": 3600
|
262 |
},
|
263 |
{
|
264 |
-
"epoch": 0.
|
265 |
-
"grad_norm": 0.
|
266 |
-
"learning_rate":
|
267 |
-
"loss": 2.
|
268 |
"step": 3700
|
269 |
},
|
270 |
{
|
271 |
-
"epoch": 0.
|
272 |
-
"grad_norm": 0.
|
273 |
-
"learning_rate":
|
274 |
-
"loss": 2.
|
275 |
"step": 3800
|
276 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
277 |
{
|
278 |
"epoch": 1.0,
|
279 |
-
"step":
|
280 |
-
"total_flos":
|
281 |
-
"train_loss": 2.
|
282 |
-
"train_runtime":
|
283 |
-
"train_samples_per_second":
|
284 |
-
"train_steps_per_second": 3.
|
285 |
}
|
286 |
],
|
287 |
"logging_steps": 100,
|
288 |
-
"max_steps":
|
289 |
"num_input_tokens_seen": 0,
|
290 |
"num_train_epochs": 1,
|
291 |
"save_steps": 500,
|
@@ -301,7 +462,7 @@
|
|
301 |
"attributes": {}
|
302 |
}
|
303 |
},
|
304 |
-
"total_flos":
|
305 |
"train_batch_size": 2,
|
306 |
"trial_name": null,
|
307 |
"trial_params": null
|
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 6120,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.016339869281045753,
|
13 |
+
"grad_norm": 0.05770206078886986,
|
14 |
+
"learning_rate": 4.085801838610828e-07,
|
15 |
+
"loss": 2.6907,
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.032679738562091505,
|
20 |
+
"grad_norm": 0.05021829903125763,
|
21 |
+
"learning_rate": 8.171603677221656e-07,
|
22 |
+
"loss": 2.6813,
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.049019607843137254,
|
27 |
+
"grad_norm": 0.06600035727024078,
|
28 |
+
"learning_rate": 1.2257405515832485e-06,
|
29 |
+
"loss": 2.699,
|
30 |
"step": 300
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.06535947712418301,
|
34 |
+
"grad_norm": 0.07294648885726929,
|
35 |
+
"learning_rate": 1.6343207354443311e-06,
|
36 |
+
"loss": 2.6897,
|
37 |
"step": 400
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.08169934640522876,
|
41 |
+
"grad_norm": 0.08237992972135544,
|
42 |
+
"learning_rate": 2.042900919305414e-06,
|
43 |
+
"loss": 2.6484,
|
44 |
"step": 500
|
45 |
},
|
46 |
{
|
47 |
+
"epoch": 0.09803921568627451,
|
48 |
+
"grad_norm": 0.0935598686337471,
|
49 |
+
"learning_rate": 2.451481103166497e-06,
|
50 |
+
"loss": 2.6607,
|
51 |
"step": 600
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 0.11437908496732026,
|
55 |
+
"grad_norm": 0.10522934049367905,
|
56 |
+
"learning_rate": 2.8600612870275796e-06,
|
57 |
+
"loss": 2.6464,
|
58 |
"step": 700
|
59 |
},
|
60 |
{
|
61 |
+
"epoch": 0.13071895424836602,
|
62 |
+
"grad_norm": 0.12197130918502808,
|
63 |
+
"learning_rate": 3.2686414708886623e-06,
|
64 |
+
"loss": 2.6393,
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 0.14705882352941177,
|
69 |
+
"grad_norm": 0.12897726893424988,
|
70 |
+
"learning_rate": 3.6772216547497446e-06,
|
71 |
+
"loss": 2.6262,
|
72 |
"step": 900
|
73 |
},
|
74 |
{
|
75 |
+
"epoch": 0.16339869281045752,
|
76 |
+
"grad_norm": 0.1309213489294052,
|
77 |
+
"learning_rate": 4.085801838610828e-06,
|
78 |
+
"loss": 2.617,
|
79 |
"step": 1000
|
80 |
},
|
81 |
{
|
82 |
+
"epoch": 0.17973856209150327,
|
83 |
+
"grad_norm": 0.1498122215270996,
|
84 |
+
"learning_rate": 4.494382022471911e-06,
|
85 |
+
"loss": 2.6183,
|
86 |
"step": 1100
|
87 |
},
|
88 |
{
|
89 |
+
"epoch": 0.19607843137254902,
|
90 |
+
"grad_norm": 0.1503172665834427,
|
91 |
+
"learning_rate": 4.902962206332994e-06,
|
92 |
+
"loss": 2.6103,
|
93 |
"step": 1200
|
94 |
},
|
95 |
{
|
96 |
+
"epoch": 0.21241830065359477,
|
97 |
+
"grad_norm": 0.16900745034217834,
|
98 |
+
"learning_rate": 5.311542390194075e-06,
|
99 |
+
"loss": 2.5884,
|
100 |
"step": 1300
|
101 |
},
|
102 |
{
|
103 |
+
"epoch": 0.22875816993464052,
|
104 |
+
"grad_norm": 0.16944456100463867,
|
105 |
+
"learning_rate": 5.720122574055159e-06,
|
106 |
+
"loss": 2.5994,
|
107 |
"step": 1400
|
108 |
},
|
109 |
{
|
110 |
+
"epoch": 0.24509803921568626,
|
111 |
+
"grad_norm": 0.18548929691314697,
|
112 |
+
"learning_rate": 6.1287027579162415e-06,
|
113 |
+
"loss": 2.5932,
|
114 |
"step": 1500
|
115 |
},
|
116 |
{
|
117 |
+
"epoch": 0.26143790849673204,
|
118 |
+
"grad_norm": 0.2077314257621765,
|
119 |
+
"learning_rate": 6.537282941777325e-06,
|
120 |
+
"loss": 2.5718,
|
121 |
"step": 1600
|
122 |
},
|
123 |
{
|
124 |
+
"epoch": 0.2777777777777778,
|
125 |
+
"grad_norm": 0.2167440503835678,
|
126 |
+
"learning_rate": 6.945863125638407e-06,
|
127 |
+
"loss": 2.5758,
|
128 |
"step": 1700
|
129 |
},
|
130 |
{
|
131 |
+
"epoch": 0.29411764705882354,
|
132 |
+
"grad_norm": 0.21927382051944733,
|
133 |
+
"learning_rate": 7.354443309499489e-06,
|
134 |
+
"loss": 2.5607,
|
135 |
"step": 1800
|
136 |
},
|
137 |
{
|
138 |
+
"epoch": 0.3104575163398693,
|
139 |
+
"grad_norm": 0.21177807450294495,
|
140 |
+
"learning_rate": 7.763023493360572e-06,
|
141 |
+
"loss": 2.5597,
|
142 |
"step": 1900
|
143 |
},
|
144 |
{
|
145 |
+
"epoch": 0.32679738562091504,
|
146 |
+
"grad_norm": 0.2335982471704483,
|
147 |
+
"learning_rate": 8.171603677221655e-06,
|
148 |
+
"loss": 2.5518,
|
149 |
"step": 2000
|
150 |
},
|
151 |
{
|
152 |
+
"epoch": 0.3431372549019608,
|
153 |
+
"grad_norm": 0.23431697487831116,
|
154 |
+
"learning_rate": 8.580183861082738e-06,
|
155 |
+
"loss": 2.5276,
|
156 |
"step": 2100
|
157 |
},
|
158 |
{
|
159 |
+
"epoch": 0.35947712418300654,
|
160 |
+
"grad_norm": 0.26647108793258667,
|
161 |
+
"learning_rate": 8.988764044943822e-06,
|
162 |
+
"loss": 2.5402,
|
163 |
"step": 2200
|
164 |
},
|
165 |
{
|
166 |
+
"epoch": 0.3758169934640523,
|
167 |
+
"grad_norm": 0.27895891666412354,
|
168 |
+
"learning_rate": 9.397344228804903e-06,
|
169 |
+
"loss": 2.5168,
|
170 |
"step": 2300
|
171 |
},
|
172 |
{
|
173 |
+
"epoch": 0.39215686274509803,
|
174 |
+
"grad_norm": 0.2913816273212433,
|
175 |
+
"learning_rate": 9.805924412665988e-06,
|
176 |
+
"loss": 2.5188,
|
177 |
"step": 2400
|
178 |
},
|
179 |
{
|
180 |
+
"epoch": 0.4084967320261438,
|
181 |
+
"grad_norm": 0.26464876532554626,
|
182 |
+
"learning_rate": 1.0214504596527069e-05,
|
183 |
+
"loss": 2.509,
|
184 |
"step": 2500
|
185 |
},
|
186 |
{
|
187 |
+
"epoch": 0.42483660130718953,
|
188 |
+
"grad_norm": 0.3019787669181824,
|
189 |
+
"learning_rate": 1.062308478038815e-05,
|
190 |
+
"loss": 2.5055,
|
191 |
"step": 2600
|
192 |
},
|
193 |
{
|
194 |
+
"epoch": 0.4411764705882353,
|
195 |
+
"grad_norm": 0.3169211447238922,
|
196 |
+
"learning_rate": 1.1031664964249235e-05,
|
197 |
+
"loss": 2.498,
|
198 |
"step": 2700
|
199 |
},
|
200 |
{
|
201 |
+
"epoch": 0.45751633986928103,
|
202 |
+
"grad_norm": 0.2861559987068176,
|
203 |
+
"learning_rate": 1.1440245148110318e-05,
|
204 |
+
"loss": 2.4814,
|
205 |
"step": 2800
|
206 |
},
|
207 |
{
|
208 |
+
"epoch": 0.4738562091503268,
|
209 |
+
"grad_norm": 0.30439117550849915,
|
210 |
+
"learning_rate": 1.1848825331971402e-05,
|
211 |
+
"loss": 2.4866,
|
212 |
"step": 2900
|
213 |
},
|
214 |
{
|
215 |
+
"epoch": 0.49019607843137253,
|
216 |
+
"grad_norm": 0.3125614821910858,
|
217 |
+
"learning_rate": 1.2257405515832483e-05,
|
218 |
+
"loss": 2.4667,
|
219 |
"step": 3000
|
220 |
},
|
221 |
{
|
222 |
+
"epoch": 0.5065359477124183,
|
223 |
+
"grad_norm": 0.33123213052749634,
|
224 |
+
"learning_rate": 1.2665985699693566e-05,
|
225 |
+
"loss": 2.4812,
|
226 |
"step": 3100
|
227 |
},
|
228 |
{
|
229 |
+
"epoch": 0.5228758169934641,
|
230 |
+
"grad_norm": 0.37793877720832825,
|
231 |
+
"learning_rate": 1.307456588355465e-05,
|
232 |
+
"loss": 2.4739,
|
233 |
"step": 3200
|
234 |
},
|
235 |
{
|
236 |
+
"epoch": 0.5392156862745098,
|
237 |
+
"grad_norm": 0.39265045523643494,
|
238 |
+
"learning_rate": 1.348314606741573e-05,
|
239 |
+
"loss": 2.4524,
|
240 |
"step": 3300
|
241 |
},
|
242 |
{
|
243 |
+
"epoch": 0.5555555555555556,
|
244 |
+
"grad_norm": 0.32753610610961914,
|
245 |
+
"learning_rate": 1.3891726251276814e-05,
|
246 |
+
"loss": 2.4668,
|
247 |
"step": 3400
|
248 |
},
|
249 |
{
|
250 |
+
"epoch": 0.5718954248366013,
|
251 |
+
"grad_norm": 0.34903815388679504,
|
252 |
+
"learning_rate": 1.4300306435137897e-05,
|
253 |
+
"loss": 2.4633,
|
254 |
"step": 3500
|
255 |
},
|
256 |
{
|
257 |
+
"epoch": 0.5882352941176471,
|
258 |
+
"grad_norm": 0.41583487391471863,
|
259 |
+
"learning_rate": 1.4708886618998978e-05,
|
260 |
+
"loss": 2.4481,
|
261 |
"step": 3600
|
262 |
},
|
263 |
{
|
264 |
+
"epoch": 0.6045751633986928,
|
265 |
+
"grad_norm": 0.36139875650405884,
|
266 |
+
"learning_rate": 1.5117466802860063e-05,
|
267 |
+
"loss": 2.4422,
|
268 |
"step": 3700
|
269 |
},
|
270 |
{
|
271 |
+
"epoch": 0.6209150326797386,
|
272 |
+
"grad_norm": 0.39948606491088867,
|
273 |
+
"learning_rate": 1.5526046986721144e-05,
|
274 |
+
"loss": 2.443,
|
275 |
"step": 3800
|
276 |
},
|
277 |
+
{
|
278 |
+
"epoch": 0.6372549019607843,
|
279 |
+
"grad_norm": 0.393020898103714,
|
280 |
+
"learning_rate": 1.593462717058223e-05,
|
281 |
+
"loss": 2.4375,
|
282 |
+
"step": 3900
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.6535947712418301,
|
286 |
+
"grad_norm": 0.39577871561050415,
|
287 |
+
"learning_rate": 1.634320735444331e-05,
|
288 |
+
"loss": 2.4391,
|
289 |
+
"step": 4000
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.6699346405228758,
|
293 |
+
"grad_norm": 0.38020357489585876,
|
294 |
+
"learning_rate": 1.6751787538304395e-05,
|
295 |
+
"loss": 2.4203,
|
296 |
+
"step": 4100
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.6862745098039216,
|
300 |
+
"grad_norm": 0.38298335671424866,
|
301 |
+
"learning_rate": 1.7160367722165477e-05,
|
302 |
+
"loss": 2.4129,
|
303 |
+
"step": 4200
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.7026143790849673,
|
307 |
+
"grad_norm": 0.4053588807582855,
|
308 |
+
"learning_rate": 1.7568947906026558e-05,
|
309 |
+
"loss": 2.4055,
|
310 |
+
"step": 4300
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.7189542483660131,
|
314 |
+
"grad_norm": 0.40465113520622253,
|
315 |
+
"learning_rate": 1.7977528089887643e-05,
|
316 |
+
"loss": 2.3975,
|
317 |
+
"step": 4400
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.7352941176470589,
|
321 |
+
"grad_norm": 0.4370296895503998,
|
322 |
+
"learning_rate": 1.8386108273748724e-05,
|
323 |
+
"loss": 2.3822,
|
324 |
+
"step": 4500
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.7516339869281046,
|
328 |
+
"grad_norm": 0.44034600257873535,
|
329 |
+
"learning_rate": 1.8794688457609806e-05,
|
330 |
+
"loss": 2.4174,
|
331 |
+
"step": 4600
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.7679738562091504,
|
335 |
+
"grad_norm": 0.45608291029930115,
|
336 |
+
"learning_rate": 1.920326864147089e-05,
|
337 |
+
"loss": 2.4016,
|
338 |
+
"step": 4700
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.7843137254901961,
|
342 |
+
"grad_norm": 0.4033947288990021,
|
343 |
+
"learning_rate": 1.9611848825331975e-05,
|
344 |
+
"loss": 2.3975,
|
345 |
+
"step": 4800
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.8006535947712419,
|
349 |
+
"grad_norm": 0.4388486444950104,
|
350 |
+
"learning_rate": 1.9999177886783194e-05,
|
351 |
+
"loss": 2.4068,
|
352 |
+
"step": 4900
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.8169934640522876,
|
356 |
+
"grad_norm": 0.4106648564338684,
|
357 |
+
"learning_rate": 1.9639628606958535e-05,
|
358 |
+
"loss": 2.3926,
|
359 |
+
"step": 5000
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.8333333333333334,
|
363 |
+
"grad_norm": 0.41734156012535095,
|
364 |
+
"learning_rate": 1.8649548579446938e-05,
|
365 |
+
"loss": 2.3915,
|
366 |
+
"step": 5100
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.8496732026143791,
|
370 |
+
"grad_norm": 0.41231608390808105,
|
371 |
+
"learning_rate": 1.709369921005258e-05,
|
372 |
+
"loss": 2.3906,
|
373 |
+
"step": 5200
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.8660130718954249,
|
377 |
+
"grad_norm": 0.4466901421546936,
|
378 |
+
"learning_rate": 1.5073849032208823e-05,
|
379 |
+
"loss": 2.383,
|
380 |
+
"step": 5300
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.8823529411764706,
|
384 |
+
"grad_norm": 0.44348859786987305,
|
385 |
+
"learning_rate": 1.2722116999329712e-05,
|
386 |
+
"loss": 2.369,
|
387 |
+
"step": 5400
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.8986928104575164,
|
391 |
+
"grad_norm": 0.4730405807495117,
|
392 |
+
"learning_rate": 1.0192330547876871e-05,
|
393 |
+
"loss": 2.3652,
|
394 |
+
"step": 5500
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.9150326797385621,
|
398 |
+
"grad_norm": 0.48740801215171814,
|
399 |
+
"learning_rate": 7.649963702603848e-06,
|
400 |
+
"loss": 2.3792,
|
401 |
+
"step": 5600
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.9313725490196079,
|
405 |
+
"grad_norm": 0.39891138672828674,
|
406 |
+
"learning_rate": 5.2613133752700145e-06,
|
407 |
+
"loss": 2.3746,
|
408 |
+
"step": 5700
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.9477124183006536,
|
412 |
+
"grad_norm": 0.45519766211509705,
|
413 |
+
"learning_rate": 3.1826218380900066e-06,
|
414 |
+
"loss": 2.3756,
|
415 |
+
"step": 5800
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.9640522875816994,
|
419 |
+
"grad_norm": 0.4656579792499542,
|
420 |
+
"learning_rate": 1.5498568739099907e-06,
|
421 |
+
"loss": 2.3776,
|
422 |
+
"step": 5900
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.9803921568627451,
|
426 |
+
"grad_norm": 0.4382289946079254,
|
427 |
+
"learning_rate": 4.6981808628712823e-07,
|
428 |
+
"loss": 2.3749,
|
429 |
+
"step": 6000
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.9967320261437909,
|
433 |
+
"grad_norm": 0.49260008335113525,
|
434 |
+
"learning_rate": 1.3151108166156168e-08,
|
435 |
+
"loss": 2.3821,
|
436 |
+
"step": 6100
|
437 |
+
},
|
438 |
{
|
439 |
"epoch": 1.0,
|
440 |
+
"step": 6120,
|
441 |
+
"total_flos": 8.917940750474281e+17,
|
442 |
+
"train_loss": 2.4952426230985356,
|
443 |
+
"train_runtime": 1729.9843,
|
444 |
+
"train_samples_per_second": 56.6,
|
445 |
+
"train_steps_per_second": 3.538
|
446 |
}
|
447 |
],
|
448 |
"logging_steps": 100,
|
449 |
+
"max_steps": 6120,
|
450 |
"num_input_tokens_seen": 0,
|
451 |
"num_train_epochs": 1,
|
452 |
"save_steps": 500,
|
|
|
462 |
"attributes": {}
|
463 |
}
|
464 |
},
|
465 |
+
"total_flos": 8.917940750474281e+17,
|
466 |
"train_batch_size": 2,
|
467 |
"trial_name": null,
|
468 |
"trial_params": null
|