sedrickkeh
commited on
Commit
•
c5f7dd4
1
Parent(s):
e4d16de
End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +836 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
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_mistral_adambeta1_0.92_dcftv1.2
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# hp_ablations_mistral_adambeta1_0.92_dcftv1.2
|
17 |
|
18 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on
|
19 |
It achieves the following results on the evaluation set:
|
20 |
- Loss: 0.0765
|
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_mistral_adambeta1_0.92_dcftv1.2
|
|
|
16 |
|
17 |
# hp_ablations_mistral_adambeta1_0.92_dcftv1.2
|
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-v1.2_no-curation_gpt-4o-mini dataset.
|
20 |
It achieves the following results on the evaluation set:
|
21 |
- Loss: 0.0765
|
22 |
|
all_results.json
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.99695843190267,
|
3 |
+
"eval_loss": 0.07649821043014526,
|
4 |
+
"eval_runtime": 384.4835,
|
5 |
+
"eval_samples_per_second": 25.918,
|
6 |
+
"eval_steps_per_second": 0.406,
|
7 |
+
"total_flos": 1854056851046400.0,
|
8 |
+
"train_loss": 0.49201587243670264,
|
9 |
+
"train_runtime": 63342.9422,
|
10 |
+
"train_samples_per_second": 8.967,
|
11 |
+
"train_steps_per_second": 0.017
|
12 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.99695843190267,
|
3 |
+
"eval_loss": 0.07649821043014526,
|
4 |
+
"eval_runtime": 384.4835,
|
5 |
+
"eval_samples_per_second": 25.918,
|
6 |
+
"eval_steps_per_second": 0.406
|
7 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.99695843190267,
|
3 |
+
"total_flos": 1854056851046400.0,
|
4 |
+
"train_loss": 0.49201587243670264,
|
5 |
+
"train_runtime": 63342.9422,
|
6 |
+
"train_samples_per_second": 8.967,
|
7 |
+
"train_steps_per_second": 0.017
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,836 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.99695843190267,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1107,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.027036160865157147,
|
13 |
+
"grad_norm": 3.3627634794530086,
|
14 |
+
"learning_rate": 5e-06,
|
15 |
+
"loss": 0.8839,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.054072321730314295,
|
20 |
+
"grad_norm": 1.9703511972746102,
|
21 |
+
"learning_rate": 5e-06,
|
22 |
+
"loss": 0.6886,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.08110848259547145,
|
27 |
+
"grad_norm": 1.895618934030067,
|
28 |
+
"learning_rate": 5e-06,
|
29 |
+
"loss": 0.649,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.10814464346062859,
|
34 |
+
"grad_norm": 3.018653004689794,
|
35 |
+
"learning_rate": 5e-06,
|
36 |
+
"loss": 0.6375,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.13518080432578575,
|
41 |
+
"grad_norm": 2.253174703767421,
|
42 |
+
"learning_rate": 5e-06,
|
43 |
+
"loss": 0.6233,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.1622169651909429,
|
48 |
+
"grad_norm": 2.5770804247981793,
|
49 |
+
"learning_rate": 5e-06,
|
50 |
+
"loss": 0.6164,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.18925312605610004,
|
55 |
+
"grad_norm": 1.9685443465813035,
|
56 |
+
"learning_rate": 5e-06,
|
57 |
+
"loss": 0.6127,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.21628928692125718,
|
62 |
+
"grad_norm": 2.1325268775254034,
|
63 |
+
"learning_rate": 5e-06,
|
64 |
+
"loss": 0.6039,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.24332544778641432,
|
69 |
+
"grad_norm": 1.4745840951249662,
|
70 |
+
"learning_rate": 5e-06,
|
71 |
+
"loss": 0.601,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.2703616086515715,
|
76 |
+
"grad_norm": 1.971048371029162,
|
77 |
+
"learning_rate": 5e-06,
|
78 |
+
"loss": 0.6032,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.29739776951672864,
|
83 |
+
"grad_norm": 1.6569004063408272,
|
84 |
+
"learning_rate": 5e-06,
|
85 |
+
"loss": 0.6012,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.3244339303818858,
|
90 |
+
"grad_norm": 2.057401724353791,
|
91 |
+
"learning_rate": 5e-06,
|
92 |
+
"loss": 0.5971,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.3514700912470429,
|
97 |
+
"grad_norm": 1.525579699451329,
|
98 |
+
"learning_rate": 5e-06,
|
99 |
+
"loss": 0.5993,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.37850625211220007,
|
104 |
+
"grad_norm": 1.60308792852181,
|
105 |
+
"learning_rate": 5e-06,
|
106 |
+
"loss": 0.5985,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.4055424129773572,
|
111 |
+
"grad_norm": 1.6398129211146073,
|
112 |
+
"learning_rate": 5e-06,
|
113 |
+
"loss": 0.5895,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.43257857384251436,
|
118 |
+
"grad_norm": 1.9310976828981994,
|
119 |
+
"learning_rate": 5e-06,
|
120 |
+
"loss": 0.5918,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.4596147347076715,
|
125 |
+
"grad_norm": 1.385584059931609,
|
126 |
+
"learning_rate": 5e-06,
|
127 |
+
"loss": 0.593,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.48665089557282865,
|
132 |
+
"grad_norm": 1.2861159852159507,
|
133 |
+
"learning_rate": 5e-06,
|
134 |
+
"loss": 0.5918,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.5136870564379858,
|
139 |
+
"grad_norm": 1.444176832285908,
|
140 |
+
"learning_rate": 5e-06,
|
141 |
+
"loss": 0.5885,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.540723217303143,
|
146 |
+
"grad_norm": 1.4862672671337698,
|
147 |
+
"learning_rate": 5e-06,
|
148 |
+
"loss": 0.5867,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.5677593781683001,
|
153 |
+
"grad_norm": 1.5075706410261351,
|
154 |
+
"learning_rate": 5e-06,
|
155 |
+
"loss": 0.5885,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.5947955390334573,
|
160 |
+
"grad_norm": 1.2766576640698686,
|
161 |
+
"learning_rate": 5e-06,
|
162 |
+
"loss": 0.5831,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.6218316998986144,
|
167 |
+
"grad_norm": 1.3449220128756942,
|
168 |
+
"learning_rate": 5e-06,
|
169 |
+
"loss": 0.5853,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.6488678607637716,
|
174 |
+
"grad_norm": 1.2820825044433828,
|
175 |
+
"learning_rate": 5e-06,
|
176 |
+
"loss": 0.5816,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.6759040216289287,
|
181 |
+
"grad_norm": 2.0436410705486927,
|
182 |
+
"learning_rate": 5e-06,
|
183 |
+
"loss": 0.575,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.7029401824940859,
|
188 |
+
"grad_norm": 1.5111257680075043,
|
189 |
+
"learning_rate": 5e-06,
|
190 |
+
"loss": 0.5781,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.729976343359243,
|
195 |
+
"grad_norm": 1.4284810364538454,
|
196 |
+
"learning_rate": 5e-06,
|
197 |
+
"loss": 0.5805,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.7570125042244001,
|
202 |
+
"grad_norm": 1.2912410275065467,
|
203 |
+
"learning_rate": 5e-06,
|
204 |
+
"loss": 0.5753,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.7840486650895573,
|
209 |
+
"grad_norm": 1.3589332122580937,
|
210 |
+
"learning_rate": 5e-06,
|
211 |
+
"loss": 0.5757,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.8110848259547144,
|
216 |
+
"grad_norm": 1.4270507706028406,
|
217 |
+
"learning_rate": 5e-06,
|
218 |
+
"loss": 0.5759,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.8381209868198716,
|
223 |
+
"grad_norm": 1.4167431698605149,
|
224 |
+
"learning_rate": 5e-06,
|
225 |
+
"loss": 0.5728,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.8651571476850287,
|
230 |
+
"grad_norm": 1.476127949628171,
|
231 |
+
"learning_rate": 5e-06,
|
232 |
+
"loss": 0.5711,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.8921933085501859,
|
237 |
+
"grad_norm": 1.3615337414773585,
|
238 |
+
"learning_rate": 5e-06,
|
239 |
+
"loss": 0.5729,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.919229469415343,
|
244 |
+
"grad_norm": 1.3530496841079478,
|
245 |
+
"learning_rate": 5e-06,
|
246 |
+
"loss": 0.5797,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.9462656302805001,
|
251 |
+
"grad_norm": 1.6939163932161898,
|
252 |
+
"learning_rate": 5e-06,
|
253 |
+
"loss": 0.5713,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.9733017911456573,
|
258 |
+
"grad_norm": 1.2503611827765622,
|
259 |
+
"learning_rate": 5e-06,
|
260 |
+
"loss": 0.5711,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.9976343359242987,
|
265 |
+
"eval_loss": 0.07110526412725449,
|
266 |
+
"eval_runtime": 379.7453,
|
267 |
+
"eval_samples_per_second": 26.241,
|
268 |
+
"eval_steps_per_second": 0.411,
|
269 |
+
"step": 369
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 1.0023656640757013,
|
273 |
+
"grad_norm": 2.951009083063046,
|
274 |
+
"learning_rate": 5e-06,
|
275 |
+
"loss": 0.5659,
|
276 |
+
"step": 370
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 1.0294018249408583,
|
280 |
+
"grad_norm": 2.275911553065513,
|
281 |
+
"learning_rate": 5e-06,
|
282 |
+
"loss": 0.4791,
|
283 |
+
"step": 380
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 1.0564379858060156,
|
287 |
+
"grad_norm": 1.9146887520712665,
|
288 |
+
"learning_rate": 5e-06,
|
289 |
+
"loss": 0.4742,
|
290 |
+
"step": 390
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 1.0834741466711728,
|
294 |
+
"grad_norm": 1.7060151561064856,
|
295 |
+
"learning_rate": 5e-06,
|
296 |
+
"loss": 0.4754,
|
297 |
+
"step": 400
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 1.1105103075363298,
|
301 |
+
"grad_norm": 1.4307333510541713,
|
302 |
+
"learning_rate": 5e-06,
|
303 |
+
"loss": 0.4723,
|
304 |
+
"step": 410
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 1.1375464684014869,
|
308 |
+
"grad_norm": 1.3914100165165024,
|
309 |
+
"learning_rate": 5e-06,
|
310 |
+
"loss": 0.4744,
|
311 |
+
"step": 420
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 1.1645826292666441,
|
315 |
+
"grad_norm": 1.7655561621577454,
|
316 |
+
"learning_rate": 5e-06,
|
317 |
+
"loss": 0.4823,
|
318 |
+
"step": 430
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 1.1916187901318014,
|
322 |
+
"grad_norm": 1.750541560120252,
|
323 |
+
"learning_rate": 5e-06,
|
324 |
+
"loss": 0.4746,
|
325 |
+
"step": 440
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 1.2186549509969584,
|
329 |
+
"grad_norm": 1.3542180298546558,
|
330 |
+
"learning_rate": 5e-06,
|
331 |
+
"loss": 0.4795,
|
332 |
+
"step": 450
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 1.2456911118621157,
|
336 |
+
"grad_norm": 1.3487709786995525,
|
337 |
+
"learning_rate": 5e-06,
|
338 |
+
"loss": 0.4811,
|
339 |
+
"step": 460
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 1.2727272727272727,
|
343 |
+
"grad_norm": 1.4890081594482487,
|
344 |
+
"learning_rate": 5e-06,
|
345 |
+
"loss": 0.4838,
|
346 |
+
"step": 470
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 1.29976343359243,
|
350 |
+
"grad_norm": 1.6710110403178111,
|
351 |
+
"learning_rate": 5e-06,
|
352 |
+
"loss": 0.4803,
|
353 |
+
"step": 480
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 1.326799594457587,
|
357 |
+
"grad_norm": 1.3946918887630642,
|
358 |
+
"learning_rate": 5e-06,
|
359 |
+
"loss": 0.4814,
|
360 |
+
"step": 490
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 1.3538357553227442,
|
364 |
+
"grad_norm": 1.349529389304425,
|
365 |
+
"learning_rate": 5e-06,
|
366 |
+
"loss": 0.4838,
|
367 |
+
"step": 500
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 1.3808719161879013,
|
371 |
+
"grad_norm": 1.7196111346917198,
|
372 |
+
"learning_rate": 5e-06,
|
373 |
+
"loss": 0.4834,
|
374 |
+
"step": 510
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 1.4079080770530585,
|
378 |
+
"grad_norm": 1.3671419685233817,
|
379 |
+
"learning_rate": 5e-06,
|
380 |
+
"loss": 0.4868,
|
381 |
+
"step": 520
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 1.4349442379182156,
|
385 |
+
"grad_norm": 1.6390806735333066,
|
386 |
+
"learning_rate": 5e-06,
|
387 |
+
"loss": 0.4797,
|
388 |
+
"step": 530
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 1.4619803987833728,
|
392 |
+
"grad_norm": 2.033732375443223,
|
393 |
+
"learning_rate": 5e-06,
|
394 |
+
"loss": 0.4838,
|
395 |
+
"step": 540
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 1.4890165596485299,
|
399 |
+
"grad_norm": 1.6780750098228978,
|
400 |
+
"learning_rate": 5e-06,
|
401 |
+
"loss": 0.4828,
|
402 |
+
"step": 550
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 1.5160527205136871,
|
406 |
+
"grad_norm": 1.4650147086655332,
|
407 |
+
"learning_rate": 5e-06,
|
408 |
+
"loss": 0.4833,
|
409 |
+
"step": 560
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 1.5430888813788441,
|
413 |
+
"grad_norm": 1.6474963909234748,
|
414 |
+
"learning_rate": 5e-06,
|
415 |
+
"loss": 0.4801,
|
416 |
+
"step": 570
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 1.5701250422440014,
|
420 |
+
"grad_norm": 1.2672048074445312,
|
421 |
+
"learning_rate": 5e-06,
|
422 |
+
"loss": 0.4821,
|
423 |
+
"step": 580
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 1.5971612031091587,
|
427 |
+
"grad_norm": 1.2455105330473952,
|
428 |
+
"learning_rate": 5e-06,
|
429 |
+
"loss": 0.4858,
|
430 |
+
"step": 590
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 1.6241973639743157,
|
434 |
+
"grad_norm": 1.3365218119955355,
|
435 |
+
"learning_rate": 5e-06,
|
436 |
+
"loss": 0.4826,
|
437 |
+
"step": 600
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 1.6512335248394727,
|
441 |
+
"grad_norm": 1.453771936609652,
|
442 |
+
"learning_rate": 5e-06,
|
443 |
+
"loss": 0.4824,
|
444 |
+
"step": 610
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 1.67826968570463,
|
448 |
+
"grad_norm": 1.3826705153584662,
|
449 |
+
"learning_rate": 5e-06,
|
450 |
+
"loss": 0.4872,
|
451 |
+
"step": 620
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 1.7053058465697872,
|
455 |
+
"grad_norm": 1.3948399069255963,
|
456 |
+
"learning_rate": 5e-06,
|
457 |
+
"loss": 0.4923,
|
458 |
+
"step": 630
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 1.7323420074349443,
|
462 |
+
"grad_norm": 1.3015399889788772,
|
463 |
+
"learning_rate": 5e-06,
|
464 |
+
"loss": 0.4897,
|
465 |
+
"step": 640
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 1.7593781683001013,
|
469 |
+
"grad_norm": 1.5649098361174691,
|
470 |
+
"learning_rate": 5e-06,
|
471 |
+
"loss": 0.4882,
|
472 |
+
"step": 650
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 1.7864143291652586,
|
476 |
+
"grad_norm": 1.4369334673977943,
|
477 |
+
"learning_rate": 5e-06,
|
478 |
+
"loss": 0.4856,
|
479 |
+
"step": 660
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 1.8134504900304158,
|
483 |
+
"grad_norm": 1.3582444869164498,
|
484 |
+
"learning_rate": 5e-06,
|
485 |
+
"loss": 0.4872,
|
486 |
+
"step": 670
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 1.8404866508955728,
|
490 |
+
"grad_norm": 1.4410245166819187,
|
491 |
+
"learning_rate": 5e-06,
|
492 |
+
"loss": 0.4902,
|
493 |
+
"step": 680
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 1.8675228117607299,
|
497 |
+
"grad_norm": 1.2401548016118424,
|
498 |
+
"learning_rate": 5e-06,
|
499 |
+
"loss": 0.4876,
|
500 |
+
"step": 690
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 1.8945589726258871,
|
504 |
+
"grad_norm": 1.3435104700539477,
|
505 |
+
"learning_rate": 5e-06,
|
506 |
+
"loss": 0.4906,
|
507 |
+
"step": 700
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 1.9215951334910444,
|
511 |
+
"grad_norm": 1.4535634930233825,
|
512 |
+
"learning_rate": 5e-06,
|
513 |
+
"loss": 0.4955,
|
514 |
+
"step": 710
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 1.9486312943562014,
|
518 |
+
"grad_norm": 1.267760090624259,
|
519 |
+
"learning_rate": 5e-06,
|
520 |
+
"loss": 0.4887,
|
521 |
+
"step": 720
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 1.9756674552213584,
|
525 |
+
"grad_norm": 1.2488776839475728,
|
526 |
+
"learning_rate": 5e-06,
|
527 |
+
"loss": 0.4845,
|
528 |
+
"step": 730
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 1.9972963839134843,
|
532 |
+
"eval_loss": 0.07172359526157379,
|
533 |
+
"eval_runtime": 380.951,
|
534 |
+
"eval_samples_per_second": 26.158,
|
535 |
+
"eval_steps_per_second": 0.41,
|
536 |
+
"step": 738
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 2.0047313281514025,
|
540 |
+
"grad_norm": 3.0892791498740326,
|
541 |
+
"learning_rate": 5e-06,
|
542 |
+
"loss": 0.4728,
|
543 |
+
"step": 740
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 2.0317674890165596,
|
547 |
+
"grad_norm": 2.0695271423861037,
|
548 |
+
"learning_rate": 5e-06,
|
549 |
+
"loss": 0.3816,
|
550 |
+
"step": 750
|
551 |
+
},
|
552 |
+
{
|
553 |
+
"epoch": 2.0588036498817166,
|
554 |
+
"grad_norm": 1.694338271769957,
|
555 |
+
"learning_rate": 5e-06,
|
556 |
+
"loss": 0.3823,
|
557 |
+
"step": 760
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 2.085839810746874,
|
561 |
+
"grad_norm": 1.492703018660841,
|
562 |
+
"learning_rate": 5e-06,
|
563 |
+
"loss": 0.3768,
|
564 |
+
"step": 770
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 2.112875971612031,
|
568 |
+
"grad_norm": 1.6552377461262013,
|
569 |
+
"learning_rate": 5e-06,
|
570 |
+
"loss": 0.3769,
|
571 |
+
"step": 780
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 2.139912132477188,
|
575 |
+
"grad_norm": 1.9420361419361558,
|
576 |
+
"learning_rate": 5e-06,
|
577 |
+
"loss": 0.3761,
|
578 |
+
"step": 790
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 2.1669482933423456,
|
582 |
+
"grad_norm": 1.5563624108859666,
|
583 |
+
"learning_rate": 5e-06,
|
584 |
+
"loss": 0.3826,
|
585 |
+
"step": 800
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 2.1939844542075027,
|
589 |
+
"grad_norm": 1.7915429380933352,
|
590 |
+
"learning_rate": 5e-06,
|
591 |
+
"loss": 0.3804,
|
592 |
+
"step": 810
|
593 |
+
},
|
594 |
+
{
|
595 |
+
"epoch": 2.2210206150726597,
|
596 |
+
"grad_norm": 1.5584683923881884,
|
597 |
+
"learning_rate": 5e-06,
|
598 |
+
"loss": 0.3816,
|
599 |
+
"step": 820
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 2.2480567759378167,
|
603 |
+
"grad_norm": 1.60097536516568,
|
604 |
+
"learning_rate": 5e-06,
|
605 |
+
"loss": 0.3844,
|
606 |
+
"step": 830
|
607 |
+
},
|
608 |
+
{
|
609 |
+
"epoch": 2.2750929368029738,
|
610 |
+
"grad_norm": 1.6548064908062865,
|
611 |
+
"learning_rate": 5e-06,
|
612 |
+
"loss": 0.3865,
|
613 |
+
"step": 840
|
614 |
+
},
|
615 |
+
{
|
616 |
+
"epoch": 2.3021290976681312,
|
617 |
+
"grad_norm": 1.7027619140998314,
|
618 |
+
"learning_rate": 5e-06,
|
619 |
+
"loss": 0.3818,
|
620 |
+
"step": 850
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"epoch": 2.3291652585332883,
|
624 |
+
"grad_norm": 1.6016849568444829,
|
625 |
+
"learning_rate": 5e-06,
|
626 |
+
"loss": 0.3847,
|
627 |
+
"step": 860
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 2.3562014193984453,
|
631 |
+
"grad_norm": 1.8796231385046944,
|
632 |
+
"learning_rate": 5e-06,
|
633 |
+
"loss": 0.39,
|
634 |
+
"step": 870
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"epoch": 2.3832375802636028,
|
638 |
+
"grad_norm": 1.5319470307978418,
|
639 |
+
"learning_rate": 5e-06,
|
640 |
+
"loss": 0.3892,
|
641 |
+
"step": 880
|
642 |
+
},
|
643 |
+
{
|
644 |
+
"epoch": 2.41027374112876,
|
645 |
+
"grad_norm": 1.7017719120193255,
|
646 |
+
"learning_rate": 5e-06,
|
647 |
+
"loss": 0.3881,
|
648 |
+
"step": 890
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 2.437309901993917,
|
652 |
+
"grad_norm": 1.5344718368107968,
|
653 |
+
"learning_rate": 5e-06,
|
654 |
+
"loss": 0.3873,
|
655 |
+
"step": 900
|
656 |
+
},
|
657 |
+
{
|
658 |
+
"epoch": 2.464346062859074,
|
659 |
+
"grad_norm": 1.6102507634771308,
|
660 |
+
"learning_rate": 5e-06,
|
661 |
+
"loss": 0.3854,
|
662 |
+
"step": 910
|
663 |
+
},
|
664 |
+
{
|
665 |
+
"epoch": 2.4913822237242313,
|
666 |
+
"grad_norm": 1.6690069949519504,
|
667 |
+
"learning_rate": 5e-06,
|
668 |
+
"loss": 0.3872,
|
669 |
+
"step": 920
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 2.5184183845893884,
|
673 |
+
"grad_norm": 1.5743935677314018,
|
674 |
+
"learning_rate": 5e-06,
|
675 |
+
"loss": 0.3867,
|
676 |
+
"step": 930
|
677 |
+
},
|
678 |
+
{
|
679 |
+
"epoch": 2.5454545454545454,
|
680 |
+
"grad_norm": 1.5720807503966818,
|
681 |
+
"learning_rate": 5e-06,
|
682 |
+
"loss": 0.394,
|
683 |
+
"step": 940
|
684 |
+
},
|
685 |
+
{
|
686 |
+
"epoch": 2.5724907063197024,
|
687 |
+
"grad_norm": 1.4596744684339498,
|
688 |
+
"learning_rate": 5e-06,
|
689 |
+
"loss": 0.3896,
|
690 |
+
"step": 950
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 2.59952686718486,
|
694 |
+
"grad_norm": 1.4774112887538513,
|
695 |
+
"learning_rate": 5e-06,
|
696 |
+
"loss": 0.3959,
|
697 |
+
"step": 960
|
698 |
+
},
|
699 |
+
{
|
700 |
+
"epoch": 2.626563028050017,
|
701 |
+
"grad_norm": 1.6927054304904465,
|
702 |
+
"learning_rate": 5e-06,
|
703 |
+
"loss": 0.3946,
|
704 |
+
"step": 970
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 2.653599188915174,
|
708 |
+
"grad_norm": 1.6990634986226298,
|
709 |
+
"learning_rate": 5e-06,
|
710 |
+
"loss": 0.399,
|
711 |
+
"step": 980
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 2.6806353497803315,
|
715 |
+
"grad_norm": 1.5811069605653503,
|
716 |
+
"learning_rate": 5e-06,
|
717 |
+
"loss": 0.3968,
|
718 |
+
"step": 990
|
719 |
+
},
|
720 |
+
{
|
721 |
+
"epoch": 2.7076715106454885,
|
722 |
+
"grad_norm": 1.929742002611046,
|
723 |
+
"learning_rate": 5e-06,
|
724 |
+
"loss": 0.3906,
|
725 |
+
"step": 1000
|
726 |
+
},
|
727 |
+
{
|
728 |
+
"epoch": 2.7347076715106455,
|
729 |
+
"grad_norm": 1.4332871535309044,
|
730 |
+
"learning_rate": 5e-06,
|
731 |
+
"loss": 0.3984,
|
732 |
+
"step": 1010
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"epoch": 2.7617438323758026,
|
736 |
+
"grad_norm": 1.6711055842838813,
|
737 |
+
"learning_rate": 5e-06,
|
738 |
+
"loss": 0.4002,
|
739 |
+
"step": 1020
|
740 |
+
},
|
741 |
+
{
|
742 |
+
"epoch": 2.7887799932409596,
|
743 |
+
"grad_norm": 1.6261611517040526,
|
744 |
+
"learning_rate": 5e-06,
|
745 |
+
"loss": 0.3984,
|
746 |
+
"step": 1030
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 2.815816154106117,
|
750 |
+
"grad_norm": 1.4326621075330992,
|
751 |
+
"learning_rate": 5e-06,
|
752 |
+
"loss": 0.3972,
|
753 |
+
"step": 1040
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 2.842852314971274,
|
757 |
+
"grad_norm": 1.4683518261050355,
|
758 |
+
"learning_rate": 5e-06,
|
759 |
+
"loss": 0.399,
|
760 |
+
"step": 1050
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"epoch": 2.869888475836431,
|
764 |
+
"grad_norm": 1.4432147424830148,
|
765 |
+
"learning_rate": 5e-06,
|
766 |
+
"loss": 0.3953,
|
767 |
+
"step": 1060
|
768 |
+
},
|
769 |
+
{
|
770 |
+
"epoch": 2.8969246367015886,
|
771 |
+
"grad_norm": 1.4795447507798194,
|
772 |
+
"learning_rate": 5e-06,
|
773 |
+
"loss": 0.4029,
|
774 |
+
"step": 1070
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 2.9239607975667457,
|
778 |
+
"grad_norm": 1.54599126728265,
|
779 |
+
"learning_rate": 5e-06,
|
780 |
+
"loss": 0.3982,
|
781 |
+
"step": 1080
|
782 |
+
},
|
783 |
+
{
|
784 |
+
"epoch": 2.9509969584319027,
|
785 |
+
"grad_norm": 1.4383101466258315,
|
786 |
+
"learning_rate": 5e-06,
|
787 |
+
"loss": 0.4018,
|
788 |
+
"step": 1090
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 2.9780331192970597,
|
792 |
+
"grad_norm": 1.3572001471611468,
|
793 |
+
"learning_rate": 5e-06,
|
794 |
+
"loss": 0.401,
|
795 |
+
"step": 1100
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 2.99695843190267,
|
799 |
+
"eval_loss": 0.07649821043014526,
|
800 |
+
"eval_runtime": 382.4679,
|
801 |
+
"eval_samples_per_second": 26.054,
|
802 |
+
"eval_steps_per_second": 0.408,
|
803 |
+
"step": 1107
|
804 |
+
},
|
805 |
+
{
|
806 |
+
"epoch": 2.99695843190267,
|
807 |
+
"step": 1107,
|
808 |
+
"total_flos": 1854056851046400.0,
|
809 |
+
"train_loss": 0.49201587243670264,
|
810 |
+
"train_runtime": 63342.9422,
|
811 |
+
"train_samples_per_second": 8.967,
|
812 |
+
"train_steps_per_second": 0.017
|
813 |
+
}
|
814 |
+
],
|
815 |
+
"logging_steps": 10,
|
816 |
+
"max_steps": 1107,
|
817 |
+
"num_input_tokens_seen": 0,
|
818 |
+
"num_train_epochs": 3,
|
819 |
+
"save_steps": 500,
|
820 |
+
"stateful_callbacks": {
|
821 |
+
"TrainerControl": {
|
822 |
+
"args": {
|
823 |
+
"should_epoch_stop": false,
|
824 |
+
"should_evaluate": false,
|
825 |
+
"should_log": false,
|
826 |
+
"should_save": true,
|
827 |
+
"should_training_stop": true
|
828 |
+
},
|
829 |
+
"attributes": {}
|
830 |
+
}
|
831 |
+
},
|
832 |
+
"total_flos": 1854056851046400.0,
|
833 |
+
"train_batch_size": 8,
|
834 |
+
"trial_name": null,
|
835 |
+
"trial_params": null
|
836 |
+
}
|
training_eval_loss.png
ADDED
training_loss.png
ADDED