Training in progress, epoch 0
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- trainer_state.json +1143 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
@@ -20,12 +20,12 @@
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
23 |
-
"v_proj",
|
24 |
-
"k_proj",
|
25 |
"q_proj",
|
26 |
-
"fc1",
|
27 |
"dense",
|
28 |
-
"fc2"
|
|
|
|
|
|
|
29 |
],
|
30 |
"task_type": "CAUSAL_LM",
|
31 |
"use_dora": false,
|
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
|
|
|
|
23 |
"q_proj",
|
|
|
24 |
"dense",
|
25 |
+
"fc2",
|
26 |
+
"fc1",
|
27 |
+
"k_proj",
|
28 |
+
"v_proj"
|
29 |
],
|
30 |
"task_type": "CAUSAL_LM",
|
31 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 94422368
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a4f871e760b818accccf6b5f51ca6d9b0db9acebe52a6b81e388ae316dd33185
|
3 |
size 94422368
|
trainer_state.json
ADDED
@@ -0,0 +1,1143 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 9.984301412872842,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1590,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.06279434850863422,
|
13 |
+
"grad_norm": 0.7715030908584595,
|
14 |
+
"learning_rate": 0.00019998048082884911,
|
15 |
+
"loss": 1.3075,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.12558869701726844,
|
20 |
+
"grad_norm": 0.7010838389396667,
|
21 |
+
"learning_rate": 0.0001999219309353572,
|
22 |
+
"loss": 1.0321,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.18838304552590268,
|
27 |
+
"grad_norm": 0.5988379120826721,
|
28 |
+
"learning_rate": 0.00019982437317643217,
|
29 |
+
"loss": 0.9725,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.25117739403453687,
|
34 |
+
"grad_norm": 0.5993463397026062,
|
35 |
+
"learning_rate": 0.00019968784563700586,
|
36 |
+
"loss": 0.9429,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.3139717425431711,
|
41 |
+
"grad_norm": 0.5684201121330261,
|
42 |
+
"learning_rate": 0.0001995124016151664,
|
43 |
+
"loss": 0.9221,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.37676609105180536,
|
48 |
+
"grad_norm": 0.5648465752601624,
|
49 |
+
"learning_rate": 0.00019929810960135172,
|
50 |
+
"loss": 0.957,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.43956043956043955,
|
55 |
+
"grad_norm": 0.5728427767753601,
|
56 |
+
"learning_rate": 0.00019904505325161158,
|
57 |
+
"loss": 0.8807,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.5023547880690737,
|
62 |
+
"grad_norm": 0.5611833333969116,
|
63 |
+
"learning_rate": 0.00019875333135495002,
|
64 |
+
"loss": 0.8894,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.565149136577708,
|
69 |
+
"grad_norm": 0.5440887808799744,
|
70 |
+
"learning_rate": 0.00019842305779475968,
|
71 |
+
"loss": 0.9077,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.6279434850863422,
|
76 |
+
"grad_norm": 0.5734382271766663,
|
77 |
+
"learning_rate": 0.00019805436150436352,
|
78 |
+
"loss": 0.872,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.6907378335949764,
|
83 |
+
"grad_norm": 0.5940066576004028,
|
84 |
+
"learning_rate": 0.00019764738641668138,
|
85 |
+
"loss": 0.8535,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.7535321821036107,
|
90 |
+
"grad_norm": 0.5004596710205078,
|
91 |
+
"learning_rate": 0.0001972022914080411,
|
92 |
+
"loss": 0.8571,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.8163265306122449,
|
97 |
+
"grad_norm": 0.49201518297195435,
|
98 |
+
"learning_rate": 0.00019671925023615573,
|
99 |
+
"loss": 0.8738,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.8791208791208791,
|
104 |
+
"grad_norm": 0.5316975116729736,
|
105 |
+
"learning_rate": 0.00019619845147229138,
|
106 |
+
"loss": 0.8571,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.9419152276295133,
|
111 |
+
"grad_norm": 0.4743081033229828,
|
112 |
+
"learning_rate": 0.00019564009842765225,
|
113 |
+
"loss": 0.8552,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.0047095761381475,
|
118 |
+
"grad_norm": 0.488436758518219,
|
119 |
+
"learning_rate": 0.0001950444090740111,
|
120 |
+
"loss": 0.8686,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.0675039246467817,
|
125 |
+
"grad_norm": 0.552605390548706,
|
126 |
+
"learning_rate": 0.0001944116159586169,
|
127 |
+
"loss": 0.7721,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.130298273155416,
|
132 |
+
"grad_norm": 0.5751229524612427,
|
133 |
+
"learning_rate": 0.0001937419661134121,
|
134 |
+
"loss": 0.7748,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.1930926216640503,
|
139 |
+
"grad_norm": 0.4798067808151245,
|
140 |
+
"learning_rate": 0.00019303572095859546,
|
141 |
+
"loss": 0.78,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.2558869701726845,
|
146 |
+
"grad_norm": 0.5503143072128296,
|
147 |
+
"learning_rate": 0.00019229315620056803,
|
148 |
+
"loss": 0.7998,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.3186813186813187,
|
153 |
+
"grad_norm": 0.5721454620361328,
|
154 |
+
"learning_rate": 0.00019151456172430183,
|
155 |
+
"loss": 0.787,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.3814756671899528,
|
160 |
+
"grad_norm": 0.5621303915977478,
|
161 |
+
"learning_rate": 0.00019070024148017375,
|
162 |
+
"loss": 0.7816,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.4442700156985873,
|
167 |
+
"grad_norm": 0.5912421941757202,
|
168 |
+
"learning_rate": 0.00018985051336530798,
|
169 |
+
"loss": 0.7806,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.5070643642072215,
|
174 |
+
"grad_norm": 0.5405421257019043,
|
175 |
+
"learning_rate": 0.00018896570909947475,
|
176 |
+
"loss": 0.7548,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.5698587127158556,
|
181 |
+
"grad_norm": 0.47901952266693115,
|
182 |
+
"learning_rate": 0.00018804617409559198,
|
183 |
+
"loss": 0.7776,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.6326530612244898,
|
188 |
+
"grad_norm": 0.5640778541564941,
|
189 |
+
"learning_rate": 0.00018709226732488215,
|
190 |
+
"loss": 0.7683,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 1.695447409733124,
|
195 |
+
"grad_norm": 0.5678220987319946,
|
196 |
+
"learning_rate": 0.00018610436117673555,
|
197 |
+
"loss": 0.7773,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 1.7582417582417582,
|
202 |
+
"grad_norm": 0.5400906801223755,
|
203 |
+
"learning_rate": 0.00018508284131333602,
|
204 |
+
"loss": 0.7797,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 1.8210361067503924,
|
209 |
+
"grad_norm": 0.5115251541137695,
|
210 |
+
"learning_rate": 0.00018402810651910442,
|
211 |
+
"loss": 0.78,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.8838304552590266,
|
216 |
+
"grad_norm": 0.5241789221763611,
|
217 |
+
"learning_rate": 0.0001829405685450202,
|
218 |
+
"loss": 0.7978,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.9466248037676608,
|
223 |
+
"grad_norm": 0.540064811706543,
|
224 |
+
"learning_rate": 0.00018182065194788025,
|
225 |
+
"loss": 0.7754,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 2.009419152276295,
|
230 |
+
"grad_norm": 0.5090541243553162,
|
231 |
+
"learning_rate": 0.0001806687939245593,
|
232 |
+
"loss": 0.7707,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 2.072213500784929,
|
237 |
+
"grad_norm": 0.6285854578018188,
|
238 |
+
"learning_rate": 0.00017948544414133534,
|
239 |
+
"loss": 0.6989,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 2.1350078492935634,
|
244 |
+
"grad_norm": 0.5556398630142212,
|
245 |
+
"learning_rate": 0.00017827106455834733,
|
246 |
+
"loss": 0.681,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 2.197802197802198,
|
251 |
+
"grad_norm": 0.6137024760246277,
|
252 |
+
"learning_rate": 0.00017702612924925376,
|
253 |
+
"loss": 0.7002,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.260596546310832,
|
258 |
+
"grad_norm": 0.5647903680801392,
|
259 |
+
"learning_rate": 0.00017575112421616202,
|
260 |
+
"loss": 0.6796,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.3233908948194664,
|
265 |
+
"grad_norm": 0.5665746927261353,
|
266 |
+
"learning_rate": 0.00017444654719990128,
|
267 |
+
"loss": 0.7247,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 2.3861852433281006,
|
272 |
+
"grad_norm": 0.5704404711723328,
|
273 |
+
"learning_rate": 0.00017311290748571275,
|
274 |
+
"loss": 0.6982,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 2.4489795918367347,
|
279 |
+
"grad_norm": 0.6268453001976013,
|
280 |
+
"learning_rate": 0.00017175072570443312,
|
281 |
+
"loss": 0.6788,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 2.511773940345369,
|
286 |
+
"grad_norm": 0.5893650054931641,
|
287 |
+
"learning_rate": 0.00017036053362924896,
|
288 |
+
"loss": 0.7071,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 2.574568288854003,
|
293 |
+
"grad_norm": 0.5590313673019409,
|
294 |
+
"learning_rate": 0.0001689428739681012,
|
295 |
+
"loss": 0.6977,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 2.6373626373626373,
|
300 |
+
"grad_norm": 0.5971296429634094,
|
301 |
+
"learning_rate": 0.00016749830015182107,
|
302 |
+
"loss": 0.7079,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.7001569858712715,
|
307 |
+
"grad_norm": 0.60088050365448,
|
308 |
+
"learning_rate": 0.00016602737611807976,
|
309 |
+
"loss": 0.6985,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 2.7629513343799057,
|
314 |
+
"grad_norm": 0.5298681855201721,
|
315 |
+
"learning_rate": 0.00016453067609123654,
|
316 |
+
"loss": 0.7271,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 2.82574568288854,
|
321 |
+
"grad_norm": 0.6201738119125366,
|
322 |
+
"learning_rate": 0.00016300878435817113,
|
323 |
+
"loss": 0.6905,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 2.8885400313971745,
|
328 |
+
"grad_norm": 0.605073869228363,
|
329 |
+
"learning_rate": 0.00016146229504018775,
|
330 |
+
"loss": 0.7071,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 2.9513343799058083,
|
335 |
+
"grad_norm": 0.5479874610900879,
|
336 |
+
"learning_rate": 0.00015989181186108002,
|
337 |
+
"loss": 0.7233,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 3.014128728414443,
|
342 |
+
"grad_norm": 0.5628566145896912,
|
343 |
+
"learning_rate": 0.0001582979479114472,
|
344 |
+
"loss": 0.6769,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 3.076923076923077,
|
349 |
+
"grad_norm": 0.6384280920028687,
|
350 |
+
"learning_rate": 0.0001566813254093538,
|
351 |
+
"loss": 0.6165,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 3.1397174254317113,
|
356 |
+
"grad_norm": 0.6049516797065735,
|
357 |
+
"learning_rate": 0.00015504257545742584,
|
358 |
+
"loss": 0.6414,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 3.2025117739403455,
|
363 |
+
"grad_norm": 0.6833707690238953,
|
364 |
+
"learning_rate": 0.0001533823377964791,
|
365 |
+
"loss": 0.6559,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 3.2653061224489797,
|
370 |
+
"grad_norm": 0.6257244944572449,
|
371 |
+
"learning_rate": 0.00015170126055577462,
|
372 |
+
"loss": 0.6465,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 3.328100470957614,
|
377 |
+
"grad_norm": 0.6548301577568054,
|
378 |
+
"learning_rate": 0.00015000000000000001,
|
379 |
+
"loss": 0.6332,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 3.390894819466248,
|
384 |
+
"grad_norm": 0.6910219788551331,
|
385 |
+
"learning_rate": 0.00014827922027307451,
|
386 |
+
"loss": 0.6331,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 3.4536891679748822,
|
391 |
+
"grad_norm": 0.5907390713691711,
|
392 |
+
"learning_rate": 0.00014653959313887813,
|
393 |
+
"loss": 0.6106,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 3.5164835164835164,
|
398 |
+
"grad_norm": 0.6401733756065369,
|
399 |
+
"learning_rate": 0.00014478179771900632,
|
400 |
+
"loss": 0.642,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 3.5792778649921506,
|
405 |
+
"grad_norm": 0.5502592325210571,
|
406 |
+
"learning_rate": 0.00014300652022765207,
|
407 |
+
"loss": 0.6392,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 3.642072213500785,
|
412 |
+
"grad_norm": 0.6283037066459656,
|
413 |
+
"learning_rate": 0.00014121445370371923,
|
414 |
+
"loss": 0.612,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 3.704866562009419,
|
419 |
+
"grad_norm": 0.6005645990371704,
|
420 |
+
"learning_rate": 0.0001394062977402717,
|
421 |
+
"loss": 0.6289,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 3.767660910518053,
|
426 |
+
"grad_norm": 0.6629378199577332,
|
427 |
+
"learning_rate": 0.00013758275821142382,
|
428 |
+
"loss": 0.6281,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 3.830455259026688,
|
433 |
+
"grad_norm": 0.6070135235786438,
|
434 |
+
"learning_rate": 0.00013574454699677893,
|
435 |
+
"loss": 0.6308,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 3.8932496075353216,
|
440 |
+
"grad_norm": 0.6706437468528748,
|
441 |
+
"learning_rate": 0.00013389238170352318,
|
442 |
+
"loss": 0.6474,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 3.956043956043956,
|
447 |
+
"grad_norm": 0.591901957988739,
|
448 |
+
"learning_rate": 0.00013202698538628376,
|
449 |
+
"loss": 0.6418,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 4.01883830455259,
|
454 |
+
"grad_norm": 0.5693519711494446,
|
455 |
+
"learning_rate": 0.0001301490862648603,
|
456 |
+
"loss": 0.6092,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 4.081632653061225,
|
461 |
+
"grad_norm": 0.6287245154380798,
|
462 |
+
"learning_rate": 0.0001282594174399399,
|
463 |
+
"loss": 0.5511,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 4.144427001569858,
|
468 |
+
"grad_norm": 0.6552153825759888,
|
469 |
+
"learning_rate": 0.00012635871660690676,
|
470 |
+
"loss": 0.5828,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 4.207221350078493,
|
475 |
+
"grad_norm": 0.664696216583252,
|
476 |
+
"learning_rate": 0.00012444772576785827,
|
477 |
+
"loss": 0.5643,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 4.270015698587127,
|
482 |
+
"grad_norm": 0.6447893381118774,
|
483 |
+
"learning_rate": 0.0001225271909419395,
|
484 |
+
"loss": 0.5883,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 4.332810047095761,
|
489 |
+
"grad_norm": 0.6885414123535156,
|
490 |
+
"learning_rate": 0.00012059786187410984,
|
491 |
+
"loss": 0.5826,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 4.395604395604396,
|
496 |
+
"grad_norm": 0.608504593372345,
|
497 |
+
"learning_rate": 0.00011866049174245491,
|
498 |
+
"loss": 0.5499,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 4.45839874411303,
|
503 |
+
"grad_norm": 0.6774938702583313,
|
504 |
+
"learning_rate": 0.00011671583686415832,
|
505 |
+
"loss": 0.5785,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 4.521193092621664,
|
510 |
+
"grad_norm": 0.6806432604789734,
|
511 |
+
"learning_rate": 0.00011476465640024814,
|
512 |
+
"loss": 0.5678,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 4.583987441130298,
|
517 |
+
"grad_norm": 0.7246624827384949,
|
518 |
+
"learning_rate": 0.00011280771205923268,
|
519 |
+
"loss": 0.5797,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 4.646781789638933,
|
524 |
+
"grad_norm": 0.6730565428733826,
|
525 |
+
"learning_rate": 0.00011084576779974257,
|
526 |
+
"loss": 0.565,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 4.7095761381475665,
|
531 |
+
"grad_norm": 0.6672644019126892,
|
532 |
+
"learning_rate": 0.00010887958953229349,
|
533 |
+
"loss": 0.5624,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 4.772370486656201,
|
538 |
+
"grad_norm": 0.7056964039802551,
|
539 |
+
"learning_rate": 0.0001069099448202878,
|
540 |
+
"loss": 0.587,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 4.835164835164835,
|
545 |
+
"grad_norm": 0.7322887182235718,
|
546 |
+
"learning_rate": 0.00010493760258037031,
|
547 |
+
"loss": 0.5712,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 4.8979591836734695,
|
552 |
+
"grad_norm": 0.7399300336837769,
|
553 |
+
"learning_rate": 0.00010296333278225599,
|
554 |
+
"loss": 0.5842,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 4.960753532182103,
|
559 |
+
"grad_norm": 0.6447321176528931,
|
560 |
+
"learning_rate": 0.00010098790614814658,
|
561 |
+
"loss": 0.559,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 5.023547880690738,
|
566 |
+
"grad_norm": 0.7017489075660706,
|
567 |
+
"learning_rate": 9.901209385185345e-05,
|
568 |
+
"loss": 0.5439,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 5.086342229199372,
|
573 |
+
"grad_norm": 0.7099691033363342,
|
574 |
+
"learning_rate": 9.703666721774402e-05,
|
575 |
+
"loss": 0.5144,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 5.149136577708006,
|
580 |
+
"grad_norm": 0.6762498617172241,
|
581 |
+
"learning_rate": 9.506239741962971e-05,
|
582 |
+
"loss": 0.5159,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 5.211930926216641,
|
587 |
+
"grad_norm": 0.722254753112793,
|
588 |
+
"learning_rate": 9.309005517971222e-05,
|
589 |
+
"loss": 0.5174,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 5.274725274725275,
|
594 |
+
"grad_norm": 0.7049859166145325,
|
595 |
+
"learning_rate": 9.112041046770653e-05,
|
596 |
+
"loss": 0.5067,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 5.337519623233909,
|
601 |
+
"grad_norm": 0.6614891886711121,
|
602 |
+
"learning_rate": 8.915423220025747e-05,
|
603 |
+
"loss": 0.5178,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 5.400313971742543,
|
608 |
+
"grad_norm": 0.7367340326309204,
|
609 |
+
"learning_rate": 8.719228794076733e-05,
|
610 |
+
"loss": 0.5158,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 5.463108320251178,
|
615 |
+
"grad_norm": 0.7544198632240295,
|
616 |
+
"learning_rate": 8.523534359975189e-05,
|
617 |
+
"loss": 0.5181,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 5.525902668759811,
|
622 |
+
"grad_norm": 0.7748332023620605,
|
623 |
+
"learning_rate": 8.328416313584169e-05,
|
624 |
+
"loss": 0.5211,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 5.588697017268446,
|
629 |
+
"grad_norm": 0.7265179753303528,
|
630 |
+
"learning_rate": 8.13395082575451e-05,
|
631 |
+
"loss": 0.5005,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 5.65149136577708,
|
636 |
+
"grad_norm": 0.7204436659812927,
|
637 |
+
"learning_rate": 7.940213812589018e-05,
|
638 |
+
"loss": 0.5034,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 5.714285714285714,
|
643 |
+
"grad_norm": 0.7397501468658447,
|
644 |
+
"learning_rate": 7.747280905806052e-05,
|
645 |
+
"loss": 0.5204,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 5.777080062794348,
|
650 |
+
"grad_norm": 0.7541511654853821,
|
651 |
+
"learning_rate": 7.555227423214174e-05,
|
652 |
+
"loss": 0.4947,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 5.839874411302983,
|
657 |
+
"grad_norm": 0.7153120636940002,
|
658 |
+
"learning_rate": 7.364128339309326e-05,
|
659 |
+
"loss": 0.5189,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 5.9026687598116165,
|
664 |
+
"grad_norm": 0.7152422070503235,
|
665 |
+
"learning_rate": 7.174058256006012e-05,
|
666 |
+
"loss": 0.5254,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 5.965463108320251,
|
671 |
+
"grad_norm": 0.7143374681472778,
|
672 |
+
"learning_rate": 6.985091373513972e-05,
|
673 |
+
"loss": 0.5044,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 6.028257456828886,
|
678 |
+
"grad_norm": 0.6690304279327393,
|
679 |
+
"learning_rate": 6.797301461371625e-05,
|
680 |
+
"loss": 0.482,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 6.0910518053375196,
|
685 |
+
"grad_norm": 0.7355020642280579,
|
686 |
+
"learning_rate": 6.610761829647685e-05,
|
687 |
+
"loss": 0.4561,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 6.153846153846154,
|
692 |
+
"grad_norm": 0.7926068902015686,
|
693 |
+
"learning_rate": 6.425545300322112e-05,
|
694 |
+
"loss": 0.4708,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 6.216640502354788,
|
699 |
+
"grad_norm": 0.7191044688224792,
|
700 |
+
"learning_rate": 6.24172417885762e-05,
|
701 |
+
"loss": 0.4586,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 6.279434850863423,
|
706 |
+
"grad_norm": 0.7463113069534302,
|
707 |
+
"learning_rate": 6.0593702259728336e-05,
|
708 |
+
"loss": 0.4674,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 6.342229199372056,
|
713 |
+
"grad_norm": 0.743307888507843,
|
714 |
+
"learning_rate": 5.8785546296280816e-05,
|
715 |
+
"loss": 0.471,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 6.405023547880691,
|
720 |
+
"grad_norm": 0.7751266956329346,
|
721 |
+
"learning_rate": 5.699347977234799e-05,
|
722 |
+
"loss": 0.4655,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 6.467817896389325,
|
727 |
+
"grad_norm": 0.8425696492195129,
|
728 |
+
"learning_rate": 5.5218202280993725e-05,
|
729 |
+
"loss": 0.476,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 6.530612244897959,
|
734 |
+
"grad_norm": 0.778333842754364,
|
735 |
+
"learning_rate": 5.3460406861121894e-05,
|
736 |
+
"loss": 0.4667,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 6.593406593406593,
|
741 |
+
"grad_norm": 0.7623078227043152,
|
742 |
+
"learning_rate": 5.172077972692553e-05,
|
743 |
+
"loss": 0.4799,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 6.656200941915228,
|
748 |
+
"grad_norm": 0.7711922526359558,
|
749 |
+
"learning_rate": 5.000000000000002e-05,
|
750 |
+
"loss": 0.4621,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 6.718995290423862,
|
755 |
+
"grad_norm": 0.812324583530426,
|
756 |
+
"learning_rate": 4.829873944422544e-05,
|
757 |
+
"loss": 0.4612,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 6.781789638932496,
|
762 |
+
"grad_norm": 0.7873876094818115,
|
763 |
+
"learning_rate": 4.661766220352097e-05,
|
764 |
+
"loss": 0.4794,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 6.84458398744113,
|
769 |
+
"grad_norm": 0.8011544942855835,
|
770 |
+
"learning_rate": 4.495742454257418e-05,
|
771 |
+
"loss": 0.4608,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 6.9073783359497645,
|
776 |
+
"grad_norm": 0.77611243724823,
|
777 |
+
"learning_rate": 4.3318674590646237e-05,
|
778 |
+
"loss": 0.471,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 6.970172684458399,
|
783 |
+
"grad_norm": 0.8160433769226074,
|
784 |
+
"learning_rate": 4.170205208855281e-05,
|
785 |
+
"loss": 0.4717,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 7.032967032967033,
|
790 |
+
"grad_norm": 0.7271689176559448,
|
791 |
+
"learning_rate": 4.010818813892e-05,
|
792 |
+
"loss": 0.4322,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 7.0957613814756675,
|
797 |
+
"grad_norm": 0.8266540765762329,
|
798 |
+
"learning_rate": 3.8537704959812294e-05,
|
799 |
+
"loss": 0.4184,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 7.158555729984301,
|
804 |
+
"grad_norm": 0.8125666975975037,
|
805 |
+
"learning_rate": 3.69912156418289e-05,
|
806 |
+
"loss": 0.4284,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 7.221350078492936,
|
811 |
+
"grad_norm": 0.839213490486145,
|
812 |
+
"learning_rate": 3.546932390876351e-05,
|
813 |
+
"loss": 0.4389,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 7.28414442700157,
|
818 |
+
"grad_norm": 0.89272141456604,
|
819 |
+
"learning_rate": 3.397262388192029e-05,
|
820 |
+
"loss": 0.4477,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 7.346938775510204,
|
825 |
+
"grad_norm": 0.7432073950767517,
|
826 |
+
"learning_rate": 3.250169984817897e-05,
|
827 |
+
"loss": 0.43,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 7.409733124018838,
|
832 |
+
"grad_norm": 0.7860047817230225,
|
833 |
+
"learning_rate": 3.105712603189884e-05,
|
834 |
+
"loss": 0.4269,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 7.472527472527473,
|
839 |
+
"grad_norm": 0.832013726234436,
|
840 |
+
"learning_rate": 2.9639466370751068e-05,
|
841 |
+
"loss": 0.4256,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 7.535321821036106,
|
846 |
+
"grad_norm": 0.7955825924873352,
|
847 |
+
"learning_rate": 2.8249274295566864e-05,
|
848 |
+
"loss": 0.4282,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 7.598116169544741,
|
853 |
+
"grad_norm": 0.7972998023033142,
|
854 |
+
"learning_rate": 2.688709251428725e-05,
|
855 |
+
"loss": 0.4388,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 7.660910518053376,
|
860 |
+
"grad_norm": 0.8105794191360474,
|
861 |
+
"learning_rate": 2.555345280009872e-05,
|
862 |
+
"loss": 0.4381,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 7.723704866562009,
|
867 |
+
"grad_norm": 0.7821062207221985,
|
868 |
+
"learning_rate": 2.4248875783837987e-05,
|
869 |
+
"loss": 0.4276,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 7.786499215070644,
|
874 |
+
"grad_norm": 0.805131733417511,
|
875 |
+
"learning_rate": 2.297387075074625e-05,
|
876 |
+
"loss": 0.4307,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 7.849293563579278,
|
881 |
+
"grad_norm": 0.797322690486908,
|
882 |
+
"learning_rate": 2.1728935441652686e-05,
|
883 |
+
"loss": 0.4295,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 7.912087912087912,
|
888 |
+
"grad_norm": 0.7943819761276245,
|
889 |
+
"learning_rate": 2.0514555858664663e-05,
|
890 |
+
"loss": 0.4346,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 7.974882260596546,
|
895 |
+
"grad_norm": 0.814616322517395,
|
896 |
+
"learning_rate": 1.93312060754407e-05,
|
897 |
+
"loss": 0.4331,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 8.03767660910518,
|
902 |
+
"grad_norm": 0.79474937915802,
|
903 |
+
"learning_rate": 1.817934805211976e-05,
|
904 |
+
"loss": 0.4153,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 8.100470957613815,
|
909 |
+
"grad_norm": 0.8072588443756104,
|
910 |
+
"learning_rate": 1.7059431454979824e-05,
|
911 |
+
"loss": 0.4072,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 8.16326530612245,
|
916 |
+
"grad_norm": 0.8833736777305603,
|
917 |
+
"learning_rate": 1.5971893480895583e-05,
|
918 |
+
"loss": 0.3996,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 8.226059654631083,
|
923 |
+
"grad_norm": 0.8510453104972839,
|
924 |
+
"learning_rate": 1.491715868666399e-05,
|
925 |
+
"loss": 0.4132,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 8.288854003139717,
|
930 |
+
"grad_norm": 0.8392331004142761,
|
931 |
+
"learning_rate": 1.3895638823264446e-05,
|
932 |
+
"loss": 0.4181,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 8.351648351648352,
|
937 |
+
"grad_norm": 0.7935534119606018,
|
938 |
+
"learning_rate": 1.290773267511788e-05,
|
939 |
+
"loss": 0.3973,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 8.414442700156986,
|
944 |
+
"grad_norm": 0.821882963180542,
|
945 |
+
"learning_rate": 1.1953825904408034e-05,
|
946 |
+
"loss": 0.4069,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 8.47723704866562,
|
951 |
+
"grad_norm": 0.8700546026229858,
|
952 |
+
"learning_rate": 1.103429090052528e-05,
|
953 |
+
"loss": 0.4144,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 8.540031397174253,
|
958 |
+
"grad_norm": 0.822404146194458,
|
959 |
+
"learning_rate": 1.0149486634692018e-05,
|
960 |
+
"loss": 0.4052,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 8.602825745682889,
|
965 |
+
"grad_norm": 0.8342509269714355,
|
966 |
+
"learning_rate": 9.299758519826273e-06,
|
967 |
+
"loss": 0.4076,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 8.665620094191523,
|
972 |
+
"grad_norm": 0.8416010141372681,
|
973 |
+
"learning_rate": 8.485438275698154e-06,
|
974 |
+
"loss": 0.4161,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 8.728414442700156,
|
979 |
+
"grad_norm": 0.7646155953407288,
|
980 |
+
"learning_rate": 7.706843799431984e-06,
|
981 |
+
"loss": 0.4153,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 8.791208791208792,
|
986 |
+
"grad_norm": 0.7787067294120789,
|
987 |
+
"learning_rate": 6.964279041404553e-06,
|
988 |
+
"loss": 0.3995,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 8.854003139717426,
|
993 |
+
"grad_norm": 0.9011712074279785,
|
994 |
+
"learning_rate": 6.258033886587911e-06,
|
995 |
+
"loss": 0.4144,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 8.91679748822606,
|
1000 |
+
"grad_norm": 0.8679221272468567,
|
1001 |
+
"learning_rate": 5.588384041383088e-06,
|
1002 |
+
"loss": 0.4147,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 8.979591836734693,
|
1007 |
+
"grad_norm": 0.8739622831344604,
|
1008 |
+
"learning_rate": 4.955590925988895e-06,
|
1009 |
+
"loss": 0.4011,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 9.042386185243329,
|
1014 |
+
"grad_norm": 0.824476420879364,
|
1015 |
+
"learning_rate": 4.359901572347758e-06,
|
1016 |
+
"loss": 0.4138,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 9.105180533751962,
|
1021 |
+
"grad_norm": 0.8952732682228088,
|
1022 |
+
"learning_rate": 3.8015485277086205e-06,
|
1023 |
+
"loss": 0.4088,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 9.167974882260596,
|
1028 |
+
"grad_norm": 0.7407020330429077,
|
1029 |
+
"learning_rate": 3.280749763844293e-06,
|
1030 |
+
"loss": 0.4018,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 9.23076923076923,
|
1035 |
+
"grad_norm": 0.8719793558120728,
|
1036 |
+
"learning_rate": 2.7977085919589254e-06,
|
1037 |
+
"loss": 0.3925,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 9.293563579277865,
|
1042 |
+
"grad_norm": 0.8144113421440125,
|
1043 |
+
"learning_rate": 2.3526135833186525e-06,
|
1044 |
+
"loss": 0.4012,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 9.3563579277865,
|
1049 |
+
"grad_norm": 0.8647878170013428,
|
1050 |
+
"learning_rate": 1.945638495636515e-06,
|
1051 |
+
"loss": 0.4006,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 9.419152276295133,
|
1056 |
+
"grad_norm": 0.827995240688324,
|
1057 |
+
"learning_rate": 1.576942205240317e-06,
|
1058 |
+
"loss": 0.3984,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 9.481946624803768,
|
1063 |
+
"grad_norm": 0.8608419895172119,
|
1064 |
+
"learning_rate": 1.2466686450499865e-06,
|
1065 |
+
"loss": 0.3956,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 9.544740973312402,
|
1070 |
+
"grad_norm": 0.8144489526748657,
|
1071 |
+
"learning_rate": 9.549467483884412e-07,
|
1072 |
+
"loss": 0.3967,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 9.607535321821036,
|
1077 |
+
"grad_norm": 0.906897783279419,
|
1078 |
+
"learning_rate": 7.018903986483083e-07,
|
1079 |
+
"loss": 0.4039,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 9.67032967032967,
|
1084 |
+
"grad_norm": 0.8140434622764587,
|
1085 |
+
"learning_rate": 4.875983848335874e-07,
|
1086 |
+
"loss": 0.3924,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 9.733124018838305,
|
1091 |
+
"grad_norm": 0.8026575446128845,
|
1092 |
+
"learning_rate": 3.12154362994177e-07,
|
1093 |
+
"loss": 0.4028,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 9.795918367346939,
|
1098 |
+
"grad_norm": 0.8668120503425598,
|
1099 |
+
"learning_rate": 1.7562682356786487e-07,
|
1100 |
+
"loss": 0.404,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 9.858712715855573,
|
1105 |
+
"grad_norm": 0.8131958842277527,
|
1106 |
+
"learning_rate": 7.806906464281615e-08,
|
1107 |
+
"loss": 0.4016,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 9.921507064364206,
|
1112 |
+
"grad_norm": 0.820686399936676,
|
1113 |
+
"learning_rate": 1.951917115091684e-08,
|
1114 |
+
"loss": 0.3993,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 9.984301412872842,
|
1119 |
+
"grad_norm": 0.8288590312004089,
|
1120 |
+
"learning_rate": 0.0,
|
1121 |
+
"loss": 0.389,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 9.984301412872842,
|
1126 |
+
"step": 1590,
|
1127 |
+
"total_flos": 1.04418647212032e+17,
|
1128 |
+
"train_loss": 0.5835858842861728,
|
1129 |
+
"train_runtime": 7264.6263,
|
1130 |
+
"train_samples_per_second": 0.877,
|
1131 |
+
"train_steps_per_second": 0.219
|
1132 |
+
}
|
1133 |
+
],
|
1134 |
+
"logging_steps": 10,
|
1135 |
+
"max_steps": 1590,
|
1136 |
+
"num_input_tokens_seen": 0,
|
1137 |
+
"num_train_epochs": 10,
|
1138 |
+
"save_steps": 500,
|
1139 |
+
"total_flos": 1.04418647212032e+17,
|
1140 |
+
"train_batch_size": 1,
|
1141 |
+
"trial_name": null,
|
1142 |
+
"trial_params": null
|
1143 |
+
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5048
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ed23d266a49157a783528fbb16fe0e775da26f9bb44ecaa76c3da7fd4d166f31
|
3 |
size 5048
|