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 56662456
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8df10efe165bcf473f61e003f2825d37c06495f93f97df26874dc4cb1e20881f
|
3 |
size 56662456
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 113488058
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c36e254af1a7713309acc3472c1cdc89752087fbcdacb0036f15ef7d8a587f38
|
3 |
size 113488058
|
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 sha256:622ac86df31aac1b538138ac38db3cc7c16c8245abf239607b9285f44e594019
|
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": 8.335,
|
75 |
"eval_steps_per_second": 4.249,
|
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": 0.2494802494802495,
|
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": 8.335,
|
75 |
"eval_steps_per_second": 4.249,
|
76 |
"step": 15
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 0.1496881496881497,
|
80 |
+
"grad_norm": 1.8020223379135132,
|
81 |
+
"learning_rate": 0.00018090169943749476,
|
82 |
+
"loss": 3.5591,
|
83 |
+
"step": 18
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 0.16632016632016633,
|
87 |
+
"eval_loss": 0.8486769199371338,
|
88 |
+
"eval_runtime": 6.1278,
|
89 |
+
"eval_samples_per_second": 8.323,
|
90 |
+
"eval_steps_per_second": 4.243,
|
91 |
+
"step": 20
|
92 |
+
},
|
93 |
+
{
|
94 |
+
"epoch": 0.17463617463617465,
|
95 |
+
"grad_norm": 1.299741506576538,
|
96 |
+
"learning_rate": 0.00013090169943749476,
|
97 |
+
"loss": 3.3536,
|
98 |
+
"step": 21
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 0.1995841995841996,
|
102 |
+
"grad_norm": 1.2542693614959717,
|
103 |
+
"learning_rate": 6.909830056250527e-05,
|
104 |
+
"loss": 3.0955,
|
105 |
+
"step": 24
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.2079002079002079,
|
109 |
+
"eval_loss": 0.8203502893447876,
|
110 |
+
"eval_runtime": 6.1584,
|
111 |
+
"eval_samples_per_second": 8.281,
|
112 |
+
"eval_steps_per_second": 4.222,
|
113 |
+
"step": 25
|
114 |
+
},
|
115 |
+
{
|
116 |
+
"epoch": 0.22453222453222454,
|
117 |
+
"grad_norm": 1.325659990310669,
|
118 |
+
"learning_rate": 1.9098300562505266e-05,
|
119 |
+
"loss": 3.2469,
|
120 |
+
"step": 27
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 0.2494802494802495,
|
124 |
+
"grad_norm": 1.3941266536712646,
|
125 |
+
"learning_rate": 0.0,
|
126 |
+
"loss": 2.9713,
|
127 |
+
"step": 30
|
128 |
+
},
|
129 |
+
{
|
130 |
+
"epoch": 0.2494802494802495,
|
131 |
+
"eval_loss": 0.8135699033737183,
|
132 |
+
"eval_runtime": 6.1516,
|
133 |
+
"eval_samples_per_second": 8.29,
|
134 |
+
"eval_steps_per_second": 4.227,
|
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": 1832860080144384.0,
|
156 |
"train_batch_size": 2,
|
157 |
"trial_name": null,
|
158 |
"trial_params": null
|