Training in progress, step 30, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 39256456
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:34887ed707a5f88d6293c9dbe624cb8bca3fecc4d1f96b52c3e5c54d98bd5a92
|
3 |
size 39256456
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 78658234
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:551933dbafc7bcd37c5145671b3a1bc42d73abfd722cb415717f61ab80314365
|
3 |
size 78658234
|
last-checkpoint/rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14244
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:577c84dcecaa557013fa5226157f60159cd3c6edb03d6b26f1df977c508647b0
|
3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1064
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2def2cd24154d8cecbaa07c36ae27e5ebb9b7273a78abfea27aa67c480e4ae2b
|
3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 0.
|
5 |
"eval_steps": 5,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -74,6 +74,65 @@
|
|
74 |
"eval_samples_per_second": 14.314,
|
75 |
"eval_steps_per_second": 7.168,
|
76 |
"step": 15
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
}
|
78 |
],
|
79 |
"logging_steps": 3,
|
@@ -88,12 +147,12 @@
|
|
88 |
"should_evaluate": false,
|
89 |
"should_log": false,
|
90 |
"should_save": true,
|
91 |
-
"should_training_stop":
|
92 |
},
|
93 |
"attributes": {}
|
94 |
}
|
95 |
},
|
96 |
-
"total_flos":
|
97 |
"train_batch_size": 2,
|
98 |
"trial_name": null,
|
99 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.019062748212867357,
|
5 |
"eval_steps": 5,
|
6 |
+
"global_step": 30,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
74 |
"eval_samples_per_second": 14.314,
|
75 |
"eval_steps_per_second": 7.168,
|
76 |
"step": 15
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 0.011437648927720413,
|
80 |
+
"grad_norm": 1.0547869205474854,
|
81 |
+
"learning_rate": 0.00018090169943749476,
|
82 |
+
"loss": 1.2018,
|
83 |
+
"step": 18
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 0.012708498808578236,
|
87 |
+
"eval_loss": 1.188161849975586,
|
88 |
+
"eval_runtime": 46.328,
|
89 |
+
"eval_samples_per_second": 14.311,
|
90 |
+
"eval_steps_per_second": 7.166,
|
91 |
+
"step": 20
|
92 |
+
},
|
93 |
+
{
|
94 |
+
"epoch": 0.013343923749007149,
|
95 |
+
"grad_norm": 0.924247145652771,
|
96 |
+
"learning_rate": 0.00013090169943749476,
|
97 |
+
"loss": 1.1364,
|
98 |
+
"step": 21
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 0.015250198570293885,
|
102 |
+
"grad_norm": 0.8752704858779907,
|
103 |
+
"learning_rate": 6.909830056250527e-05,
|
104 |
+
"loss": 1.0043,
|
105 |
+
"step": 24
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.015885623510722795,
|
109 |
+
"eval_loss": 1.125961422920227,
|
110 |
+
"eval_runtime": 46.3271,
|
111 |
+
"eval_samples_per_second": 14.311,
|
112 |
+
"eval_steps_per_second": 7.166,
|
113 |
+
"step": 25
|
114 |
+
},
|
115 |
+
{
|
116 |
+
"epoch": 0.01715647339158062,
|
117 |
+
"grad_norm": 0.8125863671302795,
|
118 |
+
"learning_rate": 1.9098300562505266e-05,
|
119 |
+
"loss": 1.0368,
|
120 |
+
"step": 27
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 0.019062748212867357,
|
124 |
+
"grad_norm": 0.8220617175102234,
|
125 |
+
"learning_rate": 0.0,
|
126 |
+
"loss": 1.0301,
|
127 |
+
"step": 30
|
128 |
+
},
|
129 |
+
{
|
130 |
+
"epoch": 0.019062748212867357,
|
131 |
+
"eval_loss": 1.1136105060577393,
|
132 |
+
"eval_runtime": 46.3216,
|
133 |
+
"eval_samples_per_second": 14.313,
|
134 |
+
"eval_steps_per_second": 7.167,
|
135 |
+
"step": 30
|
136 |
}
|
137 |
],
|
138 |
"logging_steps": 3,
|
|
|
147 |
"should_evaluate": false,
|
148 |
"should_log": false,
|
149 |
"should_save": true,
|
150 |
+
"should_training_stop": true
|
151 |
},
|
152 |
"attributes": {}
|
153 |
}
|
154 |
},
|
155 |
+
"total_flos": 5628994623897600.0,
|
156 |
"train_batch_size": 2,
|
157 |
"trial_name": null,
|
158 |
"trial_params": null
|