Upload 8 files
Browse files- config.json +2 -2
- model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +1 -1
- scheduler.pt +3 -0
- trainer_state.json +579 -43
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -13,9 +13,9 @@
|
|
| 13 |
"max_position_embeddings": 2048,
|
| 14 |
"mlp_bias": false,
|
| 15 |
"model_type": "llama",
|
| 16 |
-
"num_attention_heads":
|
| 17 |
"num_hidden_layers": 8,
|
| 18 |
-
"num_key_value_heads":
|
| 19 |
"pretraining_tp": 1,
|
| 20 |
"rms_norm_eps": 1e-06,
|
| 21 |
"rope_scaling": null,
|
|
|
|
| 13 |
"max_position_embeddings": 2048,
|
| 14 |
"mlp_bias": false,
|
| 15 |
"model_type": "llama",
|
| 16 |
+
"num_attention_heads": 16,
|
| 17 |
"num_hidden_layers": 8,
|
| 18 |
+
"num_key_value_heads": 16,
|
| 19 |
"pretraining_tp": 1,
|
| 20 |
"rms_norm_eps": 1e-06,
|
| 21 |
"rope_scaling": null,
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 18494040
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d3c2e20e14ad149e73575ba64b05ad353625b797ea427a99990d68008d3d3cb
|
| 3 |
size 18494040
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9bda737f67c64070b80a0458ecda2fcea611bd2d8344185d36fb69035c97d221
|
| 3 |
+
size 37035002
|
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 sha256:1ff264f99d31b522cc7e2a4eac9d38606d0c58a34c0adc74d71e0ca8b371dc36
|
| 3 |
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1af92ed44d240f5ab8a8e4f68304482a678ec63e16e1ccdf7c4a711760c469a9
|
| 3 |
+
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,121 +1,657 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.018807008745259066,
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"num_input_tokens_seen": 19660800,
|
| 17 |
"step": 150
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.03761401749051813,
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss":
|
| 24 |
"num_input_tokens_seen": 39321600,
|
| 25 |
"step": 300
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.0564210262357772,
|
| 29 |
-
"grad_norm": 0.
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss":
|
| 32 |
"num_input_tokens_seen": 58982400,
|
| 33 |
"step": 450
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.07522803498103627,
|
| 37 |
-
"grad_norm": 0.
|
| 38 |
-
"learning_rate":
|
| 39 |
-
"loss":
|
| 40 |
"num_input_tokens_seen": 78643200,
|
| 41 |
"step": 600
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 0.09403504372629533,
|
| 45 |
-
"grad_norm": 0.
|
| 46 |
-
"learning_rate":
|
| 47 |
-
"loss":
|
| 48 |
"num_input_tokens_seen": 98304000,
|
| 49 |
"step": 750
|
| 50 |
},
|
| 51 |
{
|
| 52 |
"epoch": 0.1128420524715544,
|
| 53 |
-
"grad_norm": 0.
|
| 54 |
-
"learning_rate":
|
| 55 |
-
"loss":
|
| 56 |
"num_input_tokens_seen": 117964800,
|
| 57 |
"step": 900
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.13164906121681347,
|
| 61 |
-
"grad_norm": 0.
|
| 62 |
-
"learning_rate":
|
| 63 |
-
"loss":
|
| 64 |
"num_input_tokens_seen": 137625600,
|
| 65 |
"step": 1050
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.15045606996207253,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss":
|
| 72 |
"num_input_tokens_seen": 157286400,
|
| 73 |
"step": 1200
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.1692630787073316,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss":
|
| 80 |
"num_input_tokens_seen": 176947200,
|
| 81 |
"step": 1350
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.18807008745259066,
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate":
|
| 87 |
-
"loss":
|
| 88 |
"num_input_tokens_seen": 196608000,
|
| 89 |
"step": 1500
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 0.20687709619784972,
|
| 93 |
-
"grad_norm":
|
| 94 |
-
"learning_rate":
|
| 95 |
-
"loss":
|
| 96 |
"num_input_tokens_seen": 216268800,
|
| 97 |
"step": 1650
|
| 98 |
},
|
| 99 |
{
|
| 100 |
"epoch": 0.2256841049431088,
|
| 101 |
-
"grad_norm":
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss":
|
| 104 |
"num_input_tokens_seen": 235929600,
|
| 105 |
"step": 1800
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.24449111368836787,
|
| 109 |
-
"grad_norm":
|
| 110 |
-
"learning_rate":
|
| 111 |
-
"loss":
|
| 112 |
"num_input_tokens_seen": 255590400,
|
| 113 |
"step": 1950
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
}
|
| 115 |
],
|
| 116 |
"logging_steps": 150,
|
| 117 |
"max_steps": 23925,
|
| 118 |
-
"num_input_tokens_seen":
|
| 119 |
"num_train_epochs": 3,
|
| 120 |
"save_steps": 500,
|
| 121 |
"stateful_callbacks": {
|
|
@@ -130,7 +666,7 @@
|
|
| 130 |
"attributes": {}
|
| 131 |
}
|
| 132 |
},
|
| 133 |
-
"total_flos":
|
| 134 |
"train_batch_size": 32,
|
| 135 |
"trial_name": null,
|
| 136 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.5045606996207255,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 12000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.018807008745259066,
|
| 13 |
+
"grad_norm": 0.8045865297317505,
|
| 14 |
+
"learning_rate": 0.00015902345219681452,
|
| 15 |
+
"loss": 9.3828,
|
| 16 |
"num_input_tokens_seen": 19660800,
|
| 17 |
"step": 150
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.03761401749051813,
|
| 21 |
+
"grad_norm": 0.8349803686141968,
|
| 22 |
+
"learning_rate": 0.0001580201496592952,
|
| 23 |
+
"loss": 8.0829,
|
| 24 |
"num_input_tokens_seen": 39321600,
|
| 25 |
"step": 300
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.0564210262357772,
|
| 29 |
+
"grad_norm": 0.6253471374511719,
|
| 30 |
+
"learning_rate": 0.00015701684712177586,
|
| 31 |
+
"loss": 7.4091,
|
| 32 |
"num_input_tokens_seen": 58982400,
|
| 33 |
"step": 450
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.07522803498103627,
|
| 37 |
+
"grad_norm": 0.5699061751365662,
|
| 38 |
+
"learning_rate": 0.00015601354458425653,
|
| 39 |
+
"loss": 7.0279,
|
| 40 |
"num_input_tokens_seen": 78643200,
|
| 41 |
"step": 600
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 0.09403504372629533,
|
| 45 |
+
"grad_norm": 0.6017518043518066,
|
| 46 |
+
"learning_rate": 0.0001550102420467372,
|
| 47 |
+
"loss": 6.732,
|
| 48 |
"num_input_tokens_seen": 98304000,
|
| 49 |
"step": 750
|
| 50 |
},
|
| 51 |
{
|
| 52 |
"epoch": 0.1128420524715544,
|
| 53 |
+
"grad_norm": 0.6045653223991394,
|
| 54 |
+
"learning_rate": 0.00015400693950921786,
|
| 55 |
+
"loss": 6.4612,
|
| 56 |
"num_input_tokens_seen": 117964800,
|
| 57 |
"step": 900
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.13164906121681347,
|
| 61 |
+
"grad_norm": 0.8820130825042725,
|
| 62 |
+
"learning_rate": 0.00015300363697169853,
|
| 63 |
+
"loss": 6.2193,
|
| 64 |
"num_input_tokens_seen": 137625600,
|
| 65 |
"step": 1050
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.15045606996207253,
|
| 69 |
+
"grad_norm": 0.8934310674667358,
|
| 70 |
+
"learning_rate": 0.00015200033443417917,
|
| 71 |
+
"loss": 6.0184,
|
| 72 |
"num_input_tokens_seen": 157286400,
|
| 73 |
"step": 1200
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.1692630787073316,
|
| 77 |
+
"grad_norm": 1.082834243774414,
|
| 78 |
+
"learning_rate": 0.00015099703189665986,
|
| 79 |
+
"loss": 5.8664,
|
| 80 |
"num_input_tokens_seen": 176947200,
|
| 81 |
"step": 1350
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.18807008745259066,
|
| 85 |
+
"grad_norm": 1.1755659580230713,
|
| 86 |
+
"learning_rate": 0.00014999372935914053,
|
| 87 |
+
"loss": 5.7258,
|
| 88 |
"num_input_tokens_seen": 196608000,
|
| 89 |
"step": 1500
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 0.20687709619784972,
|
| 93 |
+
"grad_norm": 1.1867622137069702,
|
| 94 |
+
"learning_rate": 0.00014899042682162117,
|
| 95 |
+
"loss": 5.6018,
|
| 96 |
"num_input_tokens_seen": 216268800,
|
| 97 |
"step": 1650
|
| 98 |
},
|
| 99 |
{
|
| 100 |
"epoch": 0.2256841049431088,
|
| 101 |
+
"grad_norm": 1.2599449157714844,
|
| 102 |
+
"learning_rate": 0.00014798712428410184,
|
| 103 |
+
"loss": 5.5016,
|
| 104 |
"num_input_tokens_seen": 235929600,
|
| 105 |
"step": 1800
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.24449111368836787,
|
| 109 |
+
"grad_norm": 1.292974591255188,
|
| 110 |
+
"learning_rate": 0.0001469838217465825,
|
| 111 |
+
"loss": 5.413,
|
| 112 |
"num_input_tokens_seen": 255590400,
|
| 113 |
"step": 1950
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.26329812243362694,
|
| 117 |
+
"grad_norm": 1.325179100036621,
|
| 118 |
+
"learning_rate": 0.00014598051920906318,
|
| 119 |
+
"loss": 5.3436,
|
| 120 |
+
"num_input_tokens_seen": 275251200,
|
| 121 |
+
"step": 2100
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.28210513117888597,
|
| 125 |
+
"grad_norm": 1.3993710279464722,
|
| 126 |
+
"learning_rate": 0.00014497721667154384,
|
| 127 |
+
"loss": 5.2768,
|
| 128 |
+
"num_input_tokens_seen": 294912000,
|
| 129 |
+
"step": 2250
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.30091213992414506,
|
| 133 |
+
"grad_norm": 1.5274803638458252,
|
| 134 |
+
"learning_rate": 0.0001439739141340245,
|
| 135 |
+
"loss": 5.2122,
|
| 136 |
+
"num_input_tokens_seen": 314572800,
|
| 137 |
+
"step": 2400
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.31971914866940415,
|
| 141 |
+
"grad_norm": 1.475855827331543,
|
| 142 |
+
"learning_rate": 0.00014297061159650518,
|
| 143 |
+
"loss": 5.1714,
|
| 144 |
+
"num_input_tokens_seen": 334233600,
|
| 145 |
+
"step": 2550
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.3385261574146632,
|
| 149 |
+
"grad_norm": 1.5435775518417358,
|
| 150 |
+
"learning_rate": 0.00014196730905898585,
|
| 151 |
+
"loss": 5.1111,
|
| 152 |
+
"num_input_tokens_seen": 353894400,
|
| 153 |
+
"step": 2700
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.3573331661599223,
|
| 157 |
+
"grad_norm": 1.6586663722991943,
|
| 158 |
+
"learning_rate": 0.0001409640065214665,
|
| 159 |
+
"loss": 5.0548,
|
| 160 |
+
"num_input_tokens_seen": 373555200,
|
| 161 |
+
"step": 2850
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.3761401749051813,
|
| 165 |
+
"grad_norm": 1.5785890817642212,
|
| 166 |
+
"learning_rate": 0.00013996070398394718,
|
| 167 |
+
"loss": 5.0055,
|
| 168 |
+
"num_input_tokens_seen": 393216000,
|
| 169 |
+
"step": 3000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.3949471836504404,
|
| 173 |
+
"grad_norm": 1.6534926891326904,
|
| 174 |
+
"learning_rate": 0.00013895740144642785,
|
| 175 |
+
"loss": 4.973,
|
| 176 |
+
"num_input_tokens_seen": 412876800,
|
| 177 |
+
"step": 3150
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.41375419239569944,
|
| 181 |
+
"grad_norm": 1.82899010181427,
|
| 182 |
+
"learning_rate": 0.0001379540989089085,
|
| 183 |
+
"loss": 4.9336,
|
| 184 |
+
"num_input_tokens_seen": 432537600,
|
| 185 |
+
"step": 3300
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.43256120114095853,
|
| 189 |
+
"grad_norm": 1.7137963771820068,
|
| 190 |
+
"learning_rate": 0.00013695079637138916,
|
| 191 |
+
"loss": 4.9158,
|
| 192 |
+
"num_input_tokens_seen": 452198400,
|
| 193 |
+
"step": 3450
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.4513682098862176,
|
| 197 |
+
"grad_norm": 2.001296043395996,
|
| 198 |
+
"learning_rate": 0.00013594749383386985,
|
| 199 |
+
"loss": 4.8633,
|
| 200 |
+
"num_input_tokens_seen": 471859200,
|
| 201 |
+
"step": 3600
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.47017521863147665,
|
| 205 |
+
"grad_norm": 1.7802987098693848,
|
| 206 |
+
"learning_rate": 0.0001349441912963505,
|
| 207 |
+
"loss": 4.8431,
|
| 208 |
+
"num_input_tokens_seen": 491520000,
|
| 209 |
+
"step": 3750
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.48898222737673575,
|
| 213 |
+
"grad_norm": 1.8677113056182861,
|
| 214 |
+
"learning_rate": 0.00013394088875883116,
|
| 215 |
+
"loss": 4.8243,
|
| 216 |
+
"num_input_tokens_seen": 511180800,
|
| 217 |
+
"step": 3900
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.5077892361219948,
|
| 221 |
+
"grad_norm": 1.9465229511260986,
|
| 222 |
+
"learning_rate": 0.00013293758622131183,
|
| 223 |
+
"loss": 4.7799,
|
| 224 |
+
"num_input_tokens_seen": 530841600,
|
| 225 |
+
"step": 4050
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.5265962448672539,
|
| 229 |
+
"grad_norm": 1.9226826429367065,
|
| 230 |
+
"learning_rate": 0.0001319342836837925,
|
| 231 |
+
"loss": 4.7513,
|
| 232 |
+
"num_input_tokens_seen": 550502400,
|
| 233 |
+
"step": 4200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.545403253612513,
|
| 237 |
+
"grad_norm": 2.018714666366577,
|
| 238 |
+
"learning_rate": 0.00013093098114627316,
|
| 239 |
+
"loss": 4.7226,
|
| 240 |
+
"num_input_tokens_seen": 570163200,
|
| 241 |
+
"step": 4350
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5642102623577719,
|
| 245 |
+
"grad_norm": 1.975203275680542,
|
| 246 |
+
"learning_rate": 0.00012992767860875383,
|
| 247 |
+
"loss": 4.7069,
|
| 248 |
+
"num_input_tokens_seen": 589824000,
|
| 249 |
+
"step": 4500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.583017271103031,
|
| 253 |
+
"grad_norm": 1.9473543167114258,
|
| 254 |
+
"learning_rate": 0.0001289243760712345,
|
| 255 |
+
"loss": 4.6781,
|
| 256 |
+
"num_input_tokens_seen": 609484800,
|
| 257 |
+
"step": 4650
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.6018242798482901,
|
| 261 |
+
"grad_norm": 1.9783662557601929,
|
| 262 |
+
"learning_rate": 0.00012792107353371517,
|
| 263 |
+
"loss": 4.653,
|
| 264 |
+
"num_input_tokens_seen": 629145600,
|
| 265 |
+
"step": 4800
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.6206312885935492,
|
| 269 |
+
"grad_norm": 2.184589147567749,
|
| 270 |
+
"learning_rate": 0.00012691777099619583,
|
| 271 |
+
"loss": 4.6403,
|
| 272 |
+
"num_input_tokens_seen": 648806400,
|
| 273 |
+
"step": 4950
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.6394382973388083,
|
| 277 |
+
"grad_norm": 2.109097480773926,
|
| 278 |
+
"learning_rate": 0.00012591446845867647,
|
| 279 |
+
"loss": 4.613,
|
| 280 |
+
"num_input_tokens_seen": 668467200,
|
| 281 |
+
"step": 5100
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.6582453060840673,
|
| 285 |
+
"grad_norm": 2.1296463012695312,
|
| 286 |
+
"learning_rate": 0.00012491116592115717,
|
| 287 |
+
"loss": 4.5984,
|
| 288 |
+
"num_input_tokens_seen": 688128000,
|
| 289 |
+
"step": 5250
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6770523148293264,
|
| 293 |
+
"grad_norm": 2.2577426433563232,
|
| 294 |
+
"learning_rate": 0.0001239078633836378,
|
| 295 |
+
"loss": 4.5816,
|
| 296 |
+
"num_input_tokens_seen": 707788800,
|
| 297 |
+
"step": 5400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6958593235745855,
|
| 301 |
+
"grad_norm": 2.186983585357666,
|
| 302 |
+
"learning_rate": 0.00012290456084611848,
|
| 303 |
+
"loss": 4.5545,
|
| 304 |
+
"num_input_tokens_seen": 727449600,
|
| 305 |
+
"step": 5550
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.7146663323198446,
|
| 309 |
+
"grad_norm": 2.1596477031707764,
|
| 310 |
+
"learning_rate": 0.00012190125830859916,
|
| 311 |
+
"loss": 4.5328,
|
| 312 |
+
"num_input_tokens_seen": 747110400,
|
| 313 |
+
"step": 5700
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.7334733410651036,
|
| 317 |
+
"grad_norm": 2.112995147705078,
|
| 318 |
+
"learning_rate": 0.00012089795577107981,
|
| 319 |
+
"loss": 4.5176,
|
| 320 |
+
"num_input_tokens_seen": 766771200,
|
| 321 |
+
"step": 5850
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.7522803498103626,
|
| 325 |
+
"grad_norm": 2.322986125946045,
|
| 326 |
+
"learning_rate": 0.00011989465323356048,
|
| 327 |
+
"loss": 4.5145,
|
| 328 |
+
"num_input_tokens_seen": 786432000,
|
| 329 |
+
"step": 6000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.7710873585556217,
|
| 333 |
+
"grad_norm": 2.2006442546844482,
|
| 334 |
+
"learning_rate": 0.00011889135069604115,
|
| 335 |
+
"loss": 4.4968,
|
| 336 |
+
"num_input_tokens_seen": 806092800,
|
| 337 |
+
"step": 6150
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.7898943673008808,
|
| 341 |
+
"grad_norm": 2.222529411315918,
|
| 342 |
+
"learning_rate": 0.0001178880481585218,
|
| 343 |
+
"loss": 4.4817,
|
| 344 |
+
"num_input_tokens_seen": 825753600,
|
| 345 |
+
"step": 6300
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8087013760461399,
|
| 349 |
+
"grad_norm": 2.286618947982788,
|
| 350 |
+
"learning_rate": 0.00011688474562100248,
|
| 351 |
+
"loss": 4.4596,
|
| 352 |
+
"num_input_tokens_seen": 845414400,
|
| 353 |
+
"step": 6450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.8275083847913989,
|
| 357 |
+
"grad_norm": 2.36089825630188,
|
| 358 |
+
"learning_rate": 0.0001158881317670666,
|
| 359 |
+
"loss": 4.4455,
|
| 360 |
+
"num_input_tokens_seen": 865075200,
|
| 361 |
+
"step": 6600
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.846315393536658,
|
| 365 |
+
"grad_norm": 2.4029579162597656,
|
| 366 |
+
"learning_rate": 0.00011488482922954726,
|
| 367 |
+
"loss": 4.4359,
|
| 368 |
+
"num_input_tokens_seen": 884736000,
|
| 369 |
+
"step": 6750
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.8651224022819171,
|
| 373 |
+
"grad_norm": 2.3895576000213623,
|
| 374 |
+
"learning_rate": 0.00011388152669202794,
|
| 375 |
+
"loss": 4.4152,
|
| 376 |
+
"num_input_tokens_seen": 904396800,
|
| 377 |
+
"step": 6900
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.8839294110271761,
|
| 381 |
+
"grad_norm": 2.2175564765930176,
|
| 382 |
+
"learning_rate": 0.0001128782241545086,
|
| 383 |
+
"loss": 4.4096,
|
| 384 |
+
"num_input_tokens_seen": 924057600,
|
| 385 |
+
"step": 7050
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.9027364197724352,
|
| 389 |
+
"grad_norm": 2.4382975101470947,
|
| 390 |
+
"learning_rate": 0.00011187492161698926,
|
| 391 |
+
"loss": 4.3976,
|
| 392 |
+
"num_input_tokens_seen": 943718400,
|
| 393 |
+
"step": 7200
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.9215434285176942,
|
| 397 |
+
"grad_norm": 2.533416986465454,
|
| 398 |
+
"learning_rate": 0.00011087161907946993,
|
| 399 |
+
"loss": 4.384,
|
| 400 |
+
"num_input_tokens_seen": 963379200,
|
| 401 |
+
"step": 7350
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9403504372629533,
|
| 405 |
+
"grad_norm": 2.362147808074951,
|
| 406 |
+
"learning_rate": 0.00010986831654195061,
|
| 407 |
+
"loss": 4.3608,
|
| 408 |
+
"num_input_tokens_seen": 983040000,
|
| 409 |
+
"step": 7500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.9591574460082124,
|
| 413 |
+
"grad_norm": 2.4153311252593994,
|
| 414 |
+
"learning_rate": 0.00010886501400443126,
|
| 415 |
+
"loss": 4.3604,
|
| 416 |
+
"num_input_tokens_seen": 1002700800,
|
| 417 |
+
"step": 7650
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.9779644547534715,
|
| 421 |
+
"grad_norm": 2.541653871536255,
|
| 422 |
+
"learning_rate": 0.00010786171146691193,
|
| 423 |
+
"loss": 4.3495,
|
| 424 |
+
"num_input_tokens_seen": 1022361600,
|
| 425 |
+
"step": 7800
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.9967714634987306,
|
| 429 |
+
"grad_norm": 2.507594108581543,
|
| 430 |
+
"learning_rate": 0.0001068584089293926,
|
| 431 |
+
"loss": 4.3303,
|
| 432 |
+
"num_input_tokens_seen": 1042022400,
|
| 433 |
+
"step": 7950
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 1.0155784722439896,
|
| 437 |
+
"grad_norm": 2.4842772483825684,
|
| 438 |
+
"learning_rate": 0.00010585510639187325,
|
| 439 |
+
"loss": 4.3045,
|
| 440 |
+
"num_input_tokens_seen": 1061654528,
|
| 441 |
+
"step": 8100
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.0343854809892488,
|
| 445 |
+
"grad_norm": 2.3852617740631104,
|
| 446 |
+
"learning_rate": 0.00010485180385435392,
|
| 447 |
+
"loss": 4.3033,
|
| 448 |
+
"num_input_tokens_seen": 1081315328,
|
| 449 |
+
"step": 8250
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.0531924897345077,
|
| 453 |
+
"grad_norm": 2.6390843391418457,
|
| 454 |
+
"learning_rate": 0.00010384850131683457,
|
| 455 |
+
"loss": 4.2792,
|
| 456 |
+
"num_input_tokens_seen": 1100976128,
|
| 457 |
+
"step": 8400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.0719994984797667,
|
| 461 |
+
"grad_norm": 2.5094408988952637,
|
| 462 |
+
"learning_rate": 0.00010284519877931526,
|
| 463 |
+
"loss": 4.2826,
|
| 464 |
+
"num_input_tokens_seen": 1120636928,
|
| 465 |
+
"step": 8550
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.090806507225026,
|
| 469 |
+
"grad_norm": 2.564629554748535,
|
| 470 |
+
"learning_rate": 0.00010184858492537939,
|
| 471 |
+
"loss": 4.2866,
|
| 472 |
+
"num_input_tokens_seen": 1140297728,
|
| 473 |
+
"step": 8700
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.109613515970285,
|
| 477 |
+
"grad_norm": 2.6148977279663086,
|
| 478 |
+
"learning_rate": 0.00010084528238786004,
|
| 479 |
+
"loss": 4.2614,
|
| 480 |
+
"num_input_tokens_seen": 1159958528,
|
| 481 |
+
"step": 8850
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.128420524715544,
|
| 485 |
+
"grad_norm": 2.5449106693267822,
|
| 486 |
+
"learning_rate": 9.984197985034071e-05,
|
| 487 |
+
"loss": 4.2622,
|
| 488 |
+
"num_input_tokens_seen": 1179619328,
|
| 489 |
+
"step": 9000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.147227533460803,
|
| 493 |
+
"grad_norm": 2.903010129928589,
|
| 494 |
+
"learning_rate": 9.883867731282138e-05,
|
| 495 |
+
"loss": 4.2427,
|
| 496 |
+
"num_input_tokens_seen": 1199280128,
|
| 497 |
+
"step": 9150
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.166034542206062,
|
| 501 |
+
"grad_norm": 2.6503586769104004,
|
| 502 |
+
"learning_rate": 9.783537477530203e-05,
|
| 503 |
+
"loss": 4.2433,
|
| 504 |
+
"num_input_tokens_seen": 1218940928,
|
| 505 |
+
"step": 9300
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.1848415509513213,
|
| 509 |
+
"grad_norm": 2.6648051738739014,
|
| 510 |
+
"learning_rate": 9.68320722377827e-05,
|
| 511 |
+
"loss": 4.223,
|
| 512 |
+
"num_input_tokens_seen": 1238601728,
|
| 513 |
+
"step": 9450
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.2036485596965802,
|
| 517 |
+
"grad_norm": 2.688499689102173,
|
| 518 |
+
"learning_rate": 9.582876970026338e-05,
|
| 519 |
+
"loss": 4.2232,
|
| 520 |
+
"num_input_tokens_seen": 1258262528,
|
| 521 |
+
"step": 9600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2224555684418394,
|
| 525 |
+
"grad_norm": 2.5007975101470947,
|
| 526 |
+
"learning_rate": 9.482546716274404e-05,
|
| 527 |
+
"loss": 4.2098,
|
| 528 |
+
"num_input_tokens_seen": 1277923328,
|
| 529 |
+
"step": 9750
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.2412625771870984,
|
| 533 |
+
"grad_norm": 2.682922124862671,
|
| 534 |
+
"learning_rate": 9.38221646252247e-05,
|
| 535 |
+
"loss": 4.2144,
|
| 536 |
+
"num_input_tokens_seen": 1297584128,
|
| 537 |
+
"step": 9900
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.2600695859323574,
|
| 541 |
+
"grad_norm": 2.683807611465454,
|
| 542 |
+
"learning_rate": 9.281886208770537e-05,
|
| 543 |
+
"loss": 4.1964,
|
| 544 |
+
"num_input_tokens_seen": 1317244928,
|
| 545 |
+
"step": 10050
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.2788765946776166,
|
| 549 |
+
"grad_norm": 2.7165095806121826,
|
| 550 |
+
"learning_rate": 9.181555955018603e-05,
|
| 551 |
+
"loss": 4.1939,
|
| 552 |
+
"num_input_tokens_seen": 1336905728,
|
| 553 |
+
"step": 10200
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 1.2976836034228756,
|
| 557 |
+
"grad_norm": 2.791292667388916,
|
| 558 |
+
"learning_rate": 9.081225701266671e-05,
|
| 559 |
+
"loss": 4.1796,
|
| 560 |
+
"num_input_tokens_seen": 1356566528,
|
| 561 |
+
"step": 10350
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 1.3164906121681348,
|
| 565 |
+
"grad_norm": 2.750148057937622,
|
| 566 |
+
"learning_rate": 8.980895447514737e-05,
|
| 567 |
+
"loss": 4.1778,
|
| 568 |
+
"num_input_tokens_seen": 1376227328,
|
| 569 |
+
"step": 10500
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.3352976209133938,
|
| 573 |
+
"grad_norm": 2.673332929611206,
|
| 574 |
+
"learning_rate": 8.880565193762803e-05,
|
| 575 |
+
"loss": 4.1669,
|
| 576 |
+
"num_input_tokens_seen": 1395888128,
|
| 577 |
+
"step": 10650
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.3541046296586527,
|
| 581 |
+
"grad_norm": 2.8640031814575195,
|
| 582 |
+
"learning_rate": 8.780903808369216e-05,
|
| 583 |
+
"loss": 4.167,
|
| 584 |
+
"num_input_tokens_seen": 1415548928,
|
| 585 |
+
"step": 10800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.3729116384039117,
|
| 589 |
+
"grad_norm": 2.737783432006836,
|
| 590 |
+
"learning_rate": 8.680573554617283e-05,
|
| 591 |
+
"loss": 4.1697,
|
| 592 |
+
"num_input_tokens_seen": 1435209728,
|
| 593 |
+
"step": 10950
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.391718647149171,
|
| 597 |
+
"grad_norm": 2.7067887783050537,
|
| 598 |
+
"learning_rate": 8.580243300865349e-05,
|
| 599 |
+
"loss": 4.1542,
|
| 600 |
+
"num_input_tokens_seen": 1454870528,
|
| 601 |
+
"step": 11100
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 1.41052565589443,
|
| 605 |
+
"grad_norm": 2.843256711959839,
|
| 606 |
+
"learning_rate": 8.479913047113415e-05,
|
| 607 |
+
"loss": 4.1566,
|
| 608 |
+
"num_input_tokens_seen": 1474531328,
|
| 609 |
+
"step": 11250
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 1.429332664639689,
|
| 613 |
+
"grad_norm": 2.783827543258667,
|
| 614 |
+
"learning_rate": 8.379582793361483e-05,
|
| 615 |
+
"loss": 4.1527,
|
| 616 |
+
"num_input_tokens_seen": 1494192128,
|
| 617 |
+
"step": 11400
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 1.448139673384948,
|
| 621 |
+
"grad_norm": 2.912080764770508,
|
| 622 |
+
"learning_rate": 8.279252539609549e-05,
|
| 623 |
+
"loss": 4.1308,
|
| 624 |
+
"num_input_tokens_seen": 1513852928,
|
| 625 |
+
"step": 11550
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.466946682130207,
|
| 629 |
+
"grad_norm": 2.7934062480926514,
|
| 630 |
+
"learning_rate": 8.178922285857616e-05,
|
| 631 |
+
"loss": 4.134,
|
| 632 |
+
"num_input_tokens_seen": 1533513728,
|
| 633 |
+
"step": 11700
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.4857536908754663,
|
| 637 |
+
"grad_norm": 3.1042609214782715,
|
| 638 |
+
"learning_rate": 8.078592032105681e-05,
|
| 639 |
+
"loss": 4.1365,
|
| 640 |
+
"num_input_tokens_seen": 1553174528,
|
| 641 |
+
"step": 11850
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.5045606996207255,
|
| 645 |
+
"grad_norm": 3.0361480712890625,
|
| 646 |
+
"learning_rate": 7.978261778353748e-05,
|
| 647 |
+
"loss": 4.1343,
|
| 648 |
+
"num_input_tokens_seen": 1572835328,
|
| 649 |
+
"step": 12000
|
| 650 |
}
|
| 651 |
],
|
| 652 |
"logging_steps": 150,
|
| 653 |
"max_steps": 23925,
|
| 654 |
+
"num_input_tokens_seen": 1572835328,
|
| 655 |
"num_train_epochs": 3,
|
| 656 |
"save_steps": 500,
|
| 657 |
"stateful_callbacks": {
|
|
|
|
| 666 |
"attributes": {}
|
| 667 |
}
|
| 668 |
},
|
| 669 |
+
"total_flos": 2.428558407892992e+16,
|
| 670 |
"train_batch_size": 32,
|
| 671 |
"trial_name": null,
|
| 672 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5112
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e1abd80795eaf5b1bdba294c57094293cff0723779ae399c08b79b9b31e5ed2
|
| 3 |
size 5112
|