mikhail-panzo commited on
Commit
4523398
1 Parent(s): 28b198a

Training in progress, step 1000, checkpoint

Browse files
checkpoint-1000/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:619909d8f05905cdd482bfdb5a3d15c08701ab82c621307ee9c63dcb0016dd44
3
  size 577789320
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8262041f13b16e05a0d04e7543c6599df4e2085df8b5748d4bdf3ff610d45ab
3
  size 577789320
checkpoint-1000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0153b6221f15cead1b8798c5ac71d5ea2a36de6fb796e353f29dd029a02712a0
3
  size 1155772233
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7502d80491d2506d2d361e2b088abc740a9b1147bed70ff1f5630e7f4119a1a
3
  size 1155772233
checkpoint-1000/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ad25c861e7b4be417740fc37745394735f9a504d2d18d9671b07aa1e7479a2a3
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5478438a1de30c854dabe597eb36bfc57fbab4a70587d21537bf0e86acc6df36
3
  size 14244
checkpoint-1000/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "best_metric": 0.4741882085800171,
3
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le3_s4000/checkpoint-500",
4
- "epoch": 21.978021978021978,
5
  "eval_steps": 500,
6
  "global_step": 1000,
7
  "is_hyper_param_search": false,
@@ -9,159 +9,159 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 1.098901098901099,
13
- "grad_norm": 2.8254339694976807,
14
  "learning_rate": 2.5e-05,
15
- "loss": 0.736,
16
  "step": 50
17
  },
18
  {
19
- "epoch": 2.197802197802198,
20
- "grad_norm": 4.995375633239746,
21
  "learning_rate": 4.9500000000000004e-05,
22
- "loss": 0.5913,
23
  "step": 100
24
  },
25
  {
26
- "epoch": 3.2967032967032965,
27
- "grad_norm": 2.352879285812378,
28
  "learning_rate": 7.45e-05,
29
- "loss": 0.5174,
30
  "step": 150
31
  },
32
  {
33
- "epoch": 4.395604395604396,
34
- "grad_norm": 4.0169596672058105,
35
  "learning_rate": 9.95e-05,
36
- "loss": 0.5275,
37
  "step": 200
38
  },
39
  {
40
- "epoch": 5.4945054945054945,
41
- "grad_norm": 5.557097434997559,
42
  "learning_rate": 0.0001245,
43
- "loss": 0.5231,
44
  "step": 250
45
  },
46
  {
47
- "epoch": 6.593406593406593,
48
- "grad_norm": 8.755484580993652,
49
  "learning_rate": 0.0001495,
50
- "loss": 0.5127,
51
  "step": 300
52
  },
53
  {
54
- "epoch": 7.6923076923076925,
55
- "grad_norm": 2.5811164379119873,
56
  "learning_rate": 0.00017449999999999999,
57
- "loss": 0.5179,
58
  "step": 350
59
  },
60
  {
61
- "epoch": 8.791208791208792,
62
- "grad_norm": 7.064955234527588,
63
  "learning_rate": 0.00019950000000000002,
64
- "loss": 0.4961,
65
  "step": 400
66
  },
67
  {
68
- "epoch": 9.89010989010989,
69
- "grad_norm": 8.559194564819336,
70
  "learning_rate": 0.0002245,
71
- "loss": 0.4976,
72
  "step": 450
73
  },
74
  {
75
- "epoch": 10.989010989010989,
76
- "grad_norm": 7.322611331939697,
77
  "learning_rate": 0.0002495,
78
- "loss": 0.502,
79
  "step": 500
80
  },
81
  {
82
- "epoch": 10.989010989010989,
83
- "eval_loss": 0.4741882085800171,
84
- "eval_runtime": 7.5067,
85
- "eval_samples_per_second": 21.448,
86
- "eval_steps_per_second": 2.798,
87
  "step": 500
88
  },
89
  {
90
- "epoch": 12.087912087912088,
91
- "grad_norm": 3.533902883529663,
92
  "learning_rate": 0.0002745,
93
- "loss": 0.5146,
94
  "step": 550
95
  },
96
  {
97
- "epoch": 13.186813186813186,
98
- "grad_norm": 9.026822090148926,
99
  "learning_rate": 0.0002995,
100
- "loss": 0.4956,
101
  "step": 600
102
  },
103
  {
104
- "epoch": 14.285714285714286,
105
- "grad_norm": 7.3526291847229,
106
  "learning_rate": 0.00032450000000000003,
107
- "loss": 0.5406,
108
  "step": 650
109
  },
110
  {
111
- "epoch": 15.384615384615385,
112
- "grad_norm": 2.889016628265381,
113
  "learning_rate": 0.0003495,
114
- "loss": 0.5153,
115
  "step": 700
116
  },
117
  {
118
- "epoch": 16.483516483516482,
119
- "grad_norm": 7.851646423339844,
120
  "learning_rate": 0.0003745,
121
- "loss": 0.5326,
122
  "step": 750
123
  },
124
  {
125
- "epoch": 17.582417582417584,
126
- "grad_norm": 2.468174934387207,
127
  "learning_rate": 0.0003995,
128
- "loss": 0.5575,
129
  "step": 800
130
  },
131
  {
132
- "epoch": 18.681318681318682,
133
- "grad_norm": 10.550352096557617,
134
  "learning_rate": 0.0004245,
135
- "loss": 0.5893,
136
  "step": 850
137
  },
138
  {
139
- "epoch": 19.78021978021978,
140
- "grad_norm": 6.15378475189209,
141
  "learning_rate": 0.00044950000000000003,
142
- "loss": 0.5769,
143
  "step": 900
144
  },
145
  {
146
- "epoch": 20.87912087912088,
147
- "grad_norm": 8.328354835510254,
148
  "learning_rate": 0.0004745,
149
- "loss": 0.541,
150
  "step": 950
151
  },
152
  {
153
- "epoch": 21.978021978021978,
154
- "grad_norm": 6.08469820022583,
155
  "learning_rate": 0.0004995,
156
- "loss": 0.5446,
157
  "step": 1000
158
  },
159
  {
160
- "epoch": 21.978021978021978,
161
- "eval_loss": 0.4899168014526367,
162
- "eval_runtime": 6.7313,
163
- "eval_samples_per_second": 23.918,
164
- "eval_steps_per_second": 3.12,
165
  "step": 1000
166
  }
167
  ],
@@ -182,7 +182,7 @@
182
  "attributes": {}
183
  }
184
  },
185
- "total_flos": 7353193286254464.0,
186
  "train_batch_size": 16,
187
  "trial_name": null,
188
  "trial_params": null
 
1
  {
2
+ "best_metric": 0.4463141858577728,
3
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le3_s4000/checkpoint-500",
4
+ "epoch": 22.22222222222222,
5
  "eval_steps": 500,
6
  "global_step": 1000,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 1.1111111111111112,
13
+ "grad_norm": 3.548590898513794,
14
  "learning_rate": 2.5e-05,
15
+ "loss": 0.7354,
16
  "step": 50
17
  },
18
  {
19
+ "epoch": 2.2222222222222223,
20
+ "grad_norm": 5.517350196838379,
21
  "learning_rate": 4.9500000000000004e-05,
22
+ "loss": 0.5892,
23
  "step": 100
24
  },
25
  {
26
+ "epoch": 3.3333333333333335,
27
+ "grad_norm": 4.143002033233643,
28
  "learning_rate": 7.45e-05,
29
+ "loss": 0.532,
30
  "step": 150
31
  },
32
  {
33
+ "epoch": 4.444444444444445,
34
+ "grad_norm": 6.414477825164795,
35
  "learning_rate": 9.95e-05,
36
+ "loss": 0.5284,
37
  "step": 200
38
  },
39
  {
40
+ "epoch": 5.555555555555555,
41
+ "grad_norm": 2.265143394470215,
42
  "learning_rate": 0.0001245,
43
+ "loss": 0.517,
44
  "step": 250
45
  },
46
  {
47
+ "epoch": 6.666666666666667,
48
+ "grad_norm": 4.954311847686768,
49
  "learning_rate": 0.0001495,
50
+ "loss": 0.5295,
51
  "step": 300
52
  },
53
  {
54
+ "epoch": 7.777777777777778,
55
+ "grad_norm": 2.5993571281433105,
56
  "learning_rate": 0.00017449999999999999,
57
+ "loss": 0.517,
58
  "step": 350
59
  },
60
  {
61
+ "epoch": 8.88888888888889,
62
+ "grad_norm": 3.27908992767334,
63
  "learning_rate": 0.00019950000000000002,
64
+ "loss": 0.4939,
65
  "step": 400
66
  },
67
  {
68
+ "epoch": 10.0,
69
+ "grad_norm": 4.22811222076416,
70
  "learning_rate": 0.0002245,
71
+ "loss": 0.5043,
72
  "step": 450
73
  },
74
  {
75
+ "epoch": 11.11111111111111,
76
+ "grad_norm": 2.7051336765289307,
77
  "learning_rate": 0.0002495,
78
+ "loss": 0.5028,
79
  "step": 500
80
  },
81
  {
82
+ "epoch": 11.11111111111111,
83
+ "eval_loss": 0.4463141858577728,
84
+ "eval_runtime": 16.8864,
85
+ "eval_samples_per_second": 9.416,
86
+ "eval_steps_per_second": 1.184,
87
  "step": 500
88
  },
89
  {
90
+ "epoch": 12.222222222222221,
91
+ "grad_norm": 3.0811643600463867,
92
  "learning_rate": 0.0002745,
93
+ "loss": 0.545,
94
  "step": 550
95
  },
96
  {
97
+ "epoch": 13.333333333333334,
98
+ "grad_norm": 7.707492828369141,
99
  "learning_rate": 0.0002995,
100
+ "loss": 0.4979,
101
  "step": 600
102
  },
103
  {
104
+ "epoch": 14.444444444444445,
105
+ "grad_norm": 5.85659122467041,
106
  "learning_rate": 0.00032450000000000003,
107
+ "loss": 0.4916,
108
  "step": 650
109
  },
110
  {
111
+ "epoch": 15.555555555555555,
112
+ "grad_norm": 8.300439834594727,
113
  "learning_rate": 0.0003495,
114
+ "loss": 0.5112,
115
  "step": 700
116
  },
117
  {
118
+ "epoch": 16.666666666666668,
119
+ "grad_norm": 10.664773941040039,
120
  "learning_rate": 0.0003745,
121
+ "loss": 0.5312,
122
  "step": 750
123
  },
124
  {
125
+ "epoch": 17.77777777777778,
126
+ "grad_norm": 6.4097161293029785,
127
  "learning_rate": 0.0003995,
128
+ "loss": 0.6087,
129
  "step": 800
130
  },
131
  {
132
+ "epoch": 18.88888888888889,
133
+ "grad_norm": 5.435739994049072,
134
  "learning_rate": 0.0004245,
135
+ "loss": 0.6127,
136
  "step": 850
137
  },
138
  {
139
+ "epoch": 20.0,
140
+ "grad_norm": 7.4649834632873535,
141
  "learning_rate": 0.00044950000000000003,
142
+ "loss": 0.6904,
143
  "step": 900
144
  },
145
  {
146
+ "epoch": 21.11111111111111,
147
+ "grad_norm": 10.81910228729248,
148
  "learning_rate": 0.0004745,
149
+ "loss": 0.6948,
150
  "step": 950
151
  },
152
  {
153
+ "epoch": 22.22222222222222,
154
+ "grad_norm": 4.4550886154174805,
155
  "learning_rate": 0.0004995,
156
+ "loss": 0.6348,
157
  "step": 1000
158
  },
159
  {
160
+ "epoch": 22.22222222222222,
161
+ "eval_loss": 0.5990382432937622,
162
+ "eval_runtime": 9.0191,
163
+ "eval_samples_per_second": 17.629,
164
+ "eval_steps_per_second": 2.218,
165
  "step": 1000
166
  }
167
  ],
 
182
  "attributes": {}
183
  }
184
  },
185
+ "total_flos": 6884217255754080.0,
186
  "train_batch_size": 16,
187
  "trial_name": null,
188
  "trial_params": null
checkpoint-1000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7108ae4bf8bd54061271f16e95ebc27c1c713cee5a07308d46dd91440b38577c
3
  size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86205dd0f5e4c109b754d1a2e96d6be52f2878290f435671b3cc191b44de8f2c
3
  size 5304