Upload folder using huggingface_hub
Browse files
slots/14/checkpoint-3358/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TwinyForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 3072,
|
| 11 |
+
"max_position_embeddings": 128,
|
| 12 |
+
"model_type": "twiny",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 1,
|
| 16 |
+
"num_key_value_heads": 3,
|
| 17 |
+
"qk_norm": true,
|
| 18 |
+
"rezero_init": 1.0,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.0.0",
|
| 22 |
+
"use_cache": false,
|
| 23 |
+
"vocab_size": 32000
|
| 24 |
+
}
|
slots/14/checkpoint-3358/trainer_state.json
ADDED
|
@@ -0,0 +1,1210 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8436435199966502,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3358,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 2.617016627215141e-05,
|
| 14 |
+
"grad_norm": 122.33340454101562,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 419.9996032714844,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0005234033254430282,
|
| 21 |
+
"grad_norm": 90.82410430908203,
|
| 22 |
+
"learning_rate": 4.999999511980259e-05,
|
| 23 |
+
"loss": 428.18197471217104,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0010468066508860564,
|
| 28 |
+
"grad_norm": 59.52641296386719,
|
| 29 |
+
"learning_rate": 4.999997943828395e-05,
|
| 30 |
+
"loss": 366.3947265625,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0015702099763290846,
|
| 35 |
+
"grad_norm": 41.61277389526367,
|
| 36 |
+
"learning_rate": 4.999995294193233e-05,
|
| 37 |
+
"loss": 325.6545166015625,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0020936133017721128,
|
| 42 |
+
"grad_norm": 46.67556381225586,
|
| 43 |
+
"learning_rate": 4.999991563075919e-05,
|
| 44 |
+
"loss": 310.3648193359375,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.002617016627215141,
|
| 49 |
+
"grad_norm": 47.65824890136719,
|
| 50 |
+
"learning_rate": 4.999986750478065e-05,
|
| 51 |
+
"loss": 303.49443359375,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.003140419952658169,
|
| 56 |
+
"grad_norm": 62.748138427734375,
|
| 57 |
+
"learning_rate": 4.999980856401755e-05,
|
| 58 |
+
"loss": 322.389990234375,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.007327646556202395,
|
| 63 |
+
"grad_norm": 48.20169448852539,
|
| 64 |
+
"learning_rate": 4.999895523943923e-05,
|
| 65 |
+
"loss": 288.82497336647725,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.008374453207088451,
|
| 70 |
+
"grad_norm": 68.66603088378906,
|
| 71 |
+
"learning_rate": 4.99986329623214e-05,
|
| 72 |
+
"loss": 293.87509765625,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.009421259857974507,
|
| 77 |
+
"grad_norm": 61.69401931762695,
|
| 78 |
+
"learning_rate": 4.999826742820622e-05,
|
| 79 |
+
"loss": 324.10361328125,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.010468066508860565,
|
| 84 |
+
"grad_norm": 56.778385162353516,
|
| 85 |
+
"learning_rate": 4.999785863772618e-05,
|
| 86 |
+
"loss": 274.875537109375,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.01151487315974662,
|
| 91 |
+
"grad_norm": 71.3816146850586,
|
| 92 |
+
"learning_rate": 4.999740659158865e-05,
|
| 93 |
+
"loss": 281.357080078125,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.012561679810632677,
|
| 98 |
+
"grad_norm": 75.08697509765625,
|
| 99 |
+
"learning_rate": 4.9996911290575844e-05,
|
| 100 |
+
"loss": 263.852783203125,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.013608486461518732,
|
| 105 |
+
"grad_norm": 53.91341781616211,
|
| 106 |
+
"learning_rate": 4.999637273554481e-05,
|
| 107 |
+
"loss": 258.546630859375,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.01465529311240479,
|
| 112 |
+
"grad_norm": 56.39162063598633,
|
| 113 |
+
"learning_rate": 4.9995790927427446e-05,
|
| 114 |
+
"loss": 270.5470703125,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.015702099763290844,
|
| 119 |
+
"grad_norm": 62.93777084350586,
|
| 120 |
+
"learning_rate": 4.99951658672305e-05,
|
| 121 |
+
"loss": 252.8383056640625,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.016748906414176902,
|
| 126 |
+
"grad_norm": 61.51578903198242,
|
| 127 |
+
"learning_rate": 4.9994497556035567e-05,
|
| 128 |
+
"loss": 287.8167724609375,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.01779571306506296,
|
| 133 |
+
"grad_norm": 62.75235366821289,
|
| 134 |
+
"learning_rate": 4.9993785994999074e-05,
|
| 135 |
+
"loss": 258.673779296875,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.018842519715949014,
|
| 140 |
+
"grad_norm": 56.2258186340332,
|
| 141 |
+
"learning_rate": 4.999303118535229e-05,
|
| 142 |
+
"loss": 252.9609375,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.01988932636683507,
|
| 147 |
+
"grad_norm": 54.998130798339844,
|
| 148 |
+
"learning_rate": 4.9992233128401314e-05,
|
| 149 |
+
"loss": 241.9411376953125,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.02093613301772113,
|
| 154 |
+
"grad_norm": 63.37522888183594,
|
| 155 |
+
"learning_rate": 4.999139182552709e-05,
|
| 156 |
+
"loss": 249.7280517578125,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.021982939668607183,
|
| 161 |
+
"grad_norm": 60.007144927978516,
|
| 162 |
+
"learning_rate": 4.9990507278185374e-05,
|
| 163 |
+
"loss": 236.941015625,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.02302974631949324,
|
| 168 |
+
"grad_norm": 66.06304168701172,
|
| 169 |
+
"learning_rate": 4.998957948790677e-05,
|
| 170 |
+
"loss": 240.55878363715277,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0240765529703793,
|
| 175 |
+
"grad_norm": 68.9736099243164,
|
| 176 |
+
"learning_rate": 4.998860845629671e-05,
|
| 177 |
+
"loss": 247.988720703125,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.025123359621265353,
|
| 182 |
+
"grad_norm": 61.50578308105469,
|
| 183 |
+
"learning_rate": 4.998759418503544e-05,
|
| 184 |
+
"loss": 233.025537109375,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.02617016627215141,
|
| 189 |
+
"grad_norm": 54.96660614013672,
|
| 190 |
+
"learning_rate": 4.998653667587801e-05,
|
| 191 |
+
"loss": 243.004736328125,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.06804234327362181,
|
| 196 |
+
"grad_norm": 50.733734130859375,
|
| 197 |
+
"learning_rate": 4.990903999473118e-05,
|
| 198 |
+
"loss": 250.31749378551137,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.07065935647645342,
|
| 203 |
+
"grad_norm": 48.06721496582031,
|
| 204 |
+
"learning_rate": 4.9901899191179905e-05,
|
| 205 |
+
"loss": 243.69541015625,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.07327636967928504,
|
| 210 |
+
"grad_norm": 54.863311767578125,
|
| 211 |
+
"learning_rate": 4.989448913413268e-05,
|
| 212 |
+
"loss": 225.0319580078125,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07589338288211664,
|
| 217 |
+
"grad_norm": 60.36458969116211,
|
| 218 |
+
"learning_rate": 4.988680990371125e-05,
|
| 219 |
+
"loss": 222.7537353515625,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.07851039608494825,
|
| 224 |
+
"grad_norm": 57.877986907958984,
|
| 225 |
+
"learning_rate": 4.987886158294783e-05,
|
| 226 |
+
"loss": 241.103515625,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.08112740928777985,
|
| 231 |
+
"grad_norm": 64.6783676147461,
|
| 232 |
+
"learning_rate": 4.987064425778417e-05,
|
| 233 |
+
"loss": 215.971142578125,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.08374442249061147,
|
| 238 |
+
"grad_norm": 50.828765869140625,
|
| 239 |
+
"learning_rate": 4.986215801707066e-05,
|
| 240 |
+
"loss": 213.7265380859375,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.08636143569344307,
|
| 245 |
+
"grad_norm": 55.68048858642578,
|
| 246 |
+
"learning_rate": 4.9853402952565383e-05,
|
| 247 |
+
"loss": 216.94306640625,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08897844889627468,
|
| 252 |
+
"grad_norm": 54.125267028808594,
|
| 253 |
+
"learning_rate": 4.984437915893305e-05,
|
| 254 |
+
"loss": 204.139990234375,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0915954620991063,
|
| 259 |
+
"grad_norm": 57.27487564086914,
|
| 260 |
+
"learning_rate": 4.9835086733744074e-05,
|
| 261 |
+
"loss": 202.9946533203125,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.0942124753019379,
|
| 266 |
+
"grad_norm": 65.67598724365234,
|
| 267 |
+
"learning_rate": 4.9825525777473424e-05,
|
| 268 |
+
"loss": 207.21923828125,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.09682948850476951,
|
| 273 |
+
"grad_norm": 71.1091537475586,
|
| 274 |
+
"learning_rate": 4.98156963934996e-05,
|
| 275 |
+
"loss": 204.4235107421875,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.09944650170760111,
|
| 280 |
+
"grad_norm": 60.16761779785156,
|
| 281 |
+
"learning_rate": 4.9805598688103496e-05,
|
| 282 |
+
"loss": 213.841748046875,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.19596246146437973,
|
| 287 |
+
"grad_norm": 268.7782897949219,
|
| 288 |
+
"learning_rate": 4.99999986487106e-05,
|
| 289 |
+
"loss": 474.80499267578125,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2009871399634664,
|
| 294 |
+
"grad_norm": 107.96406555175781,
|
| 295 |
+
"learning_rate": 4.9999459486180746e-05,
|
| 296 |
+
"loss": 528.41904296875,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.20601181846255306,
|
| 301 |
+
"grad_norm": 133.09701538085938,
|
| 302 |
+
"learning_rate": 4.9997837968095394e-05,
|
| 303 |
+
"loss": 484.614111328125,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.21103649696163973,
|
| 308 |
+
"grad_norm": 102.78153228759766,
|
| 309 |
+
"learning_rate": 4.999513551586019e-05,
|
| 310 |
+
"loss": 467.132373046875,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.21606117546072637,
|
| 315 |
+
"grad_norm": 92.18177795410156,
|
| 316 |
+
"learning_rate": 4.999135224633215e-05,
|
| 317 |
+
"loss": 463.333203125,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.22108585395981303,
|
| 322 |
+
"grad_norm": 106.00531005859375,
|
| 323 |
+
"learning_rate": 4.998648832310402e-05,
|
| 324 |
+
"loss": 452.96220703125,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2261105324588997,
|
| 329 |
+
"grad_norm": 100.12345123291016,
|
| 330 |
+
"learning_rate": 4.9980543956497236e-05,
|
| 331 |
+
"loss": 425.08818359375,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.23113521095798636,
|
| 336 |
+
"grad_norm": 96.5484619140625,
|
| 337 |
+
"learning_rate": 4.997351940355277e-05,
|
| 338 |
+
"loss": 474.501953125,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.23615988945707303,
|
| 343 |
+
"grad_norm": 104.5451889038086,
|
| 344 |
+
"learning_rate": 4.996541496802006e-05,
|
| 345 |
+
"loss": 447.30810546875,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.24118456795615967,
|
| 350 |
+
"grad_norm": 129.78697204589844,
|
| 351 |
+
"learning_rate": 4.995623100034386e-05,
|
| 352 |
+
"loss": 442.8482421875,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.24620924645524633,
|
| 357 |
+
"grad_norm": 95.0133285522461,
|
| 358 |
+
"learning_rate": 4.994596789764909e-05,
|
| 359 |
+
"loss": 433.66689453125,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.251233924954333,
|
| 364 |
+
"grad_norm": 93.045654296875,
|
| 365 |
+
"learning_rate": 4.993462610372364e-05,
|
| 366 |
+
"loss": 418.920263671875,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.25625860345341966,
|
| 371 |
+
"grad_norm": 102.6265869140625,
|
| 372 |
+
"learning_rate": 4.992220610899924e-05,
|
| 373 |
+
"loss": 394.33173828125,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.26128328195250633,
|
| 378 |
+
"grad_norm": 94.07565307617188,
|
| 379 |
+
"learning_rate": 4.990870845053017e-05,
|
| 380 |
+
"loss": 388.1531494140625,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.266307960451593,
|
| 385 |
+
"grad_norm": 105.10103607177734,
|
| 386 |
+
"learning_rate": 4.989413371197013e-05,
|
| 387 |
+
"loss": 379.9688720703125,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.27133263895067966,
|
| 392 |
+
"grad_norm": 99.91519927978516,
|
| 393 |
+
"learning_rate": 4.987848252354691e-05,
|
| 394 |
+
"loss": 388.9978271484375,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.2763573174497663,
|
| 399 |
+
"grad_norm": 93.43067169189453,
|
| 400 |
+
"learning_rate": 4.986175556203521e-05,
|
| 401 |
+
"loss": 359.740380859375,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.28138199594885294,
|
| 406 |
+
"grad_norm": 86.09584045410156,
|
| 407 |
+
"learning_rate": 4.984395355072732e-05,
|
| 408 |
+
"loss": 394.5279296875,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2864066744479396,
|
| 413 |
+
"grad_norm": 92.78561401367188,
|
| 414 |
+
"learning_rate": 4.982507725940191e-05,
|
| 415 |
+
"loss": 379.4552490234375,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.29143135294702627,
|
| 420 |
+
"grad_norm": 89.77680969238281,
|
| 421 |
+
"learning_rate": 4.980512750429068e-05,
|
| 422 |
+
"loss": 382.2197509765625,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.29645603144611293,
|
| 427 |
+
"grad_norm": 90.80536651611328,
|
| 428 |
+
"learning_rate": 4.97841051480431e-05,
|
| 429 |
+
"loss": 322.7532958984375,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3014807099451996,
|
| 434 |
+
"grad_norm": 95.0079574584961,
|
| 435 |
+
"learning_rate": 4.976201109968908e-05,
|
| 436 |
+
"loss": 383.434130859375,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.30650538844428626,
|
| 441 |
+
"grad_norm": 96.08088684082031,
|
| 442 |
+
"learning_rate": 4.973884631459971e-05,
|
| 443 |
+
"loss": 315.9826904296875,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.31153006694337293,
|
| 448 |
+
"grad_norm": 95.79961395263672,
|
| 449 |
+
"learning_rate": 4.9714611794445906e-05,
|
| 450 |
+
"loss": 300.992431640625,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3165547454424596,
|
| 455 |
+
"grad_norm": 85.15829467773438,
|
| 456 |
+
"learning_rate": 4.968930858715511e-05,
|
| 457 |
+
"loss": 345.0525390625,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.32157942394154626,
|
| 462 |
+
"grad_norm": 97.35791778564453,
|
| 463 |
+
"learning_rate": 4.966293778686598e-05,
|
| 464 |
+
"loss": 303.704541015625,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3266041024406329,
|
| 469 |
+
"grad_norm": 100.1771469116211,
|
| 470 |
+
"learning_rate": 4.963550053388106e-05,
|
| 471 |
+
"loss": 326.5158447265625,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.33162878093971954,
|
| 476 |
+
"grad_norm": 90.39754486083984,
|
| 477 |
+
"learning_rate": 4.9606998014617535e-05,
|
| 478 |
+
"loss": 307.4456787109375,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3366534594388062,
|
| 483 |
+
"grad_norm": 95.20532989501953,
|
| 484 |
+
"learning_rate": 4.957743146155581e-05,
|
| 485 |
+
"loss": 268.6821533203125,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.34167813793789287,
|
| 490 |
+
"grad_norm": 79.56786346435547,
|
| 491 |
+
"learning_rate": 4.954680215318635e-05,
|
| 492 |
+
"loss": 271.5860107421875,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.34670281643697953,
|
| 497 |
+
"grad_norm": 70.59774017333984,
|
| 498 |
+
"learning_rate": 4.951511141395432e-05,
|
| 499 |
+
"loss": 274.2177734375,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3517274949360662,
|
| 504 |
+
"grad_norm": 85.6360092163086,
|
| 505 |
+
"learning_rate": 4.948236061420229e-05,
|
| 506 |
+
"loss": 282.2251708984375,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.35675217343515286,
|
| 511 |
+
"grad_norm": 80.13080596923828,
|
| 512 |
+
"learning_rate": 4.944855117011107e-05,
|
| 513 |
+
"loss": 288.8053466796875,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.36177685193423953,
|
| 518 |
+
"grad_norm": 72.03929901123047,
|
| 519 |
+
"learning_rate": 4.941368454363839e-05,
|
| 520 |
+
"loss": 260.295849609375,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3668015304333262,
|
| 525 |
+
"grad_norm": 71.65677642822266,
|
| 526 |
+
"learning_rate": 4.937776224245573e-05,
|
| 527 |
+
"loss": 283.718701171875,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.37182620893241286,
|
| 532 |
+
"grad_norm": 79.88951110839844,
|
| 533 |
+
"learning_rate": 4.934078581988311e-05,
|
| 534 |
+
"loss": 237.0281494140625,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3768508874314995,
|
| 539 |
+
"grad_norm": 70.0428237915039,
|
| 540 |
+
"learning_rate": 4.930275687482191e-05,
|
| 541 |
+
"loss": 236.0194091796875,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.38187556593058614,
|
| 546 |
+
"grad_norm": 92.4908447265625,
|
| 547 |
+
"learning_rate": 4.926367705168577e-05,
|
| 548 |
+
"loss": 240.9611572265625,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.3869002444296728,
|
| 553 |
+
"grad_norm": 70.08905792236328,
|
| 554 |
+
"learning_rate": 4.922354804032944e-05,
|
| 555 |
+
"loss": 230.3265380859375,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.39192492292875947,
|
| 560 |
+
"grad_norm": 77.18536376953125,
|
| 561 |
+
"learning_rate": 4.9182371575975736e-05,
|
| 562 |
+
"loss": 242.0444580078125,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.39694960142784613,
|
| 567 |
+
"grad_norm": 79.4393081665039,
|
| 568 |
+
"learning_rate": 4.9140149439140506e-05,
|
| 569 |
+
"loss": 247.16416015625,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.4019742799269328,
|
| 574 |
+
"grad_norm": 70.19295501708984,
|
| 575 |
+
"learning_rate": 4.909688345555561e-05,
|
| 576 |
+
"loss": 221.3811279296875,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.40699895842601946,
|
| 581 |
+
"grad_norm": 66.46074676513672,
|
| 582 |
+
"learning_rate": 4.9052575496090016e-05,
|
| 583 |
+
"loss": 222.3531005859375,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.41202363692510613,
|
| 588 |
+
"grad_norm": 103.58501434326172,
|
| 589 |
+
"learning_rate": 4.900722747666888e-05,
|
| 590 |
+
"loss": 237.266455078125,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.4170483154241928,
|
| 595 |
+
"grad_norm": 84.74609375,
|
| 596 |
+
"learning_rate": 4.8960841358190704e-05,
|
| 597 |
+
"loss": 220.35146484375,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.42207299392327946,
|
| 602 |
+
"grad_norm": 75.79342651367188,
|
| 603 |
+
"learning_rate": 4.891341914644251e-05,
|
| 604 |
+
"loss": 207.2835693359375,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4270976724223661,
|
| 609 |
+
"grad_norm": 84.80883026123047,
|
| 610 |
+
"learning_rate": 4.886496289201318e-05,
|
| 611 |
+
"loss": 199.74576416015626,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.43212235092145274,
|
| 616 |
+
"grad_norm": 67.72750091552734,
|
| 617 |
+
"learning_rate": 4.881547469020471e-05,
|
| 618 |
+
"loss": 198.93392333984374,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4371470294205394,
|
| 623 |
+
"grad_norm": 76.4952163696289,
|
| 624 |
+
"learning_rate": 4.8764956680941676e-05,
|
| 625 |
+
"loss": 189.1037109375,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.44217170791962607,
|
| 630 |
+
"grad_norm": 78.63851928710938,
|
| 631 |
+
"learning_rate": 4.8713411048678635e-05,
|
| 632 |
+
"loss": 226.8509521484375,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.44719638641871273,
|
| 637 |
+
"grad_norm": 63.36662292480469,
|
| 638 |
+
"learning_rate": 4.866084002230573e-05,
|
| 639 |
+
"loss": 202.01885986328125,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4522210649177994,
|
| 644 |
+
"grad_norm": 82.46707916259766,
|
| 645 |
+
"learning_rate": 4.860724587505223e-05,
|
| 646 |
+
"loss": 190.03756103515624,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.45724574341688606,
|
| 651 |
+
"grad_norm": 79.09195709228516,
|
| 652 |
+
"learning_rate": 4.8552630924388335e-05,
|
| 653 |
+
"loss": 193.71884765625,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.46227042191597273,
|
| 658 |
+
"grad_norm": 72.94918060302734,
|
| 659 |
+
"learning_rate": 4.8496997531924885e-05,
|
| 660 |
+
"loss": 174.4781982421875,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4672951004150594,
|
| 665 |
+
"grad_norm": 97.54232025146484,
|
| 666 |
+
"learning_rate": 4.8440348103311276e-05,
|
| 667 |
+
"loss": 196.74876708984374,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.47231977891414606,
|
| 672 |
+
"grad_norm": 70.02765655517578,
|
| 673 |
+
"learning_rate": 4.838268508813143e-05,
|
| 674 |
+
"loss": 183.72100830078125,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.4773444574132327,
|
| 679 |
+
"grad_norm": 62.996543884277344,
|
| 680 |
+
"learning_rate": 4.8324010979797875e-05,
|
| 681 |
+
"loss": 171.5578369140625,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.48236913591231934,
|
| 686 |
+
"grad_norm": 72.1956787109375,
|
| 687 |
+
"learning_rate": 4.8264328315443914e-05,
|
| 688 |
+
"loss": 167.999951171875,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.487393814411406,
|
| 693 |
+
"grad_norm": 96.08892822265625,
|
| 694 |
+
"learning_rate": 4.820363967581394e-05,
|
| 695 |
+
"loss": 159.68211669921874,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.49241849291049267,
|
| 700 |
+
"grad_norm": 65.87628173828125,
|
| 701 |
+
"learning_rate": 4.814194768515182e-05,
|
| 702 |
+
"loss": 156.25810546875,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.49744317140957933,
|
| 707 |
+
"grad_norm": 70.53435516357422,
|
| 708 |
+
"learning_rate": 4.807925501108744e-05,
|
| 709 |
+
"loss": 158.4431640625,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.502467849908666,
|
| 714 |
+
"grad_norm": 71.5821304321289,
|
| 715 |
+
"learning_rate": 4.801556436452133e-05,
|
| 716 |
+
"loss": 157.15218505859374,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5074925284077526,
|
| 721 |
+
"grad_norm": 62.593109130859375,
|
| 722 |
+
"learning_rate": 4.795087849950747e-05,
|
| 723 |
+
"loss": 154.85384521484374,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5125172069068393,
|
| 728 |
+
"grad_norm": 70.64319610595703,
|
| 729 |
+
"learning_rate": 4.7885200213134164e-05,
|
| 730 |
+
"loss": 144.212548828125,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5175418854059259,
|
| 735 |
+
"grad_norm": 57.551979064941406,
|
| 736 |
+
"learning_rate": 4.781853234540312e-05,
|
| 737 |
+
"loss": 150.35511474609376,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5225665639050127,
|
| 742 |
+
"grad_norm": 60.21275329589844,
|
| 743 |
+
"learning_rate": 4.7750877779106666e-05,
|
| 744 |
+
"loss": 139.1343994140625,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5275912424040993,
|
| 749 |
+
"grad_norm": 74.57681274414062,
|
| 750 |
+
"learning_rate": 4.7682239439703025e-05,
|
| 751 |
+
"loss": 152.9317138671875,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.532615920903186,
|
| 756 |
+
"grad_norm": 62.081398010253906,
|
| 757 |
+
"learning_rate": 4.761262029518988e-05,
|
| 758 |
+
"loss": 138.762646484375,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5376405994022726,
|
| 763 |
+
"grad_norm": 97.3165283203125,
|
| 764 |
+
"learning_rate": 4.754202335597601e-05,
|
| 765 |
+
"loss": 132.7292236328125,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5426652779013593,
|
| 770 |
+
"grad_norm": 68.6803970336914,
|
| 771 |
+
"learning_rate": 4.7470451674751106e-05,
|
| 772 |
+
"loss": 129.36715087890624,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5476899564004459,
|
| 777 |
+
"grad_norm": 63.835575103759766,
|
| 778 |
+
"learning_rate": 4.7397908346353796e-05,
|
| 779 |
+
"loss": 127.2830078125,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5527146348995327,
|
| 784 |
+
"grad_norm": 63.419559478759766,
|
| 785 |
+
"learning_rate": 4.73243965076378e-05,
|
| 786 |
+
"loss": 129.15245361328124,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5577393133986193,
|
| 791 |
+
"grad_norm": 53.640995025634766,
|
| 792 |
+
"learning_rate": 4.724991933733629e-05,
|
| 793 |
+
"loss": 122.0046630859375,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5627639918977059,
|
| 798 |
+
"grad_norm": 59.515647888183594,
|
| 799 |
+
"learning_rate": 4.7174480055924455e-05,
|
| 800 |
+
"loss": 135.0306396484375,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5677886703967926,
|
| 805 |
+
"grad_norm": 61.889373779296875,
|
| 806 |
+
"learning_rate": 4.7098081925480216e-05,
|
| 807 |
+
"loss": 123.96072998046876,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5728133488958792,
|
| 812 |
+
"grad_norm": 52.50545120239258,
|
| 813 |
+
"learning_rate": 4.7020728249543196e-05,
|
| 814 |
+
"loss": 120.4759521484375,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5778380273949659,
|
| 819 |
+
"grad_norm": 64.7458267211914,
|
| 820 |
+
"learning_rate": 4.694242237297186e-05,
|
| 821 |
+
"loss": 131.4004150390625,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5828627058940525,
|
| 826 |
+
"grad_norm": 52.680599212646484,
|
| 827 |
+
"learning_rate": 4.686316768179889e-05,
|
| 828 |
+
"loss": 124.689404296875,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.5878873843931393,
|
| 833 |
+
"grad_norm": 77.16497039794922,
|
| 834 |
+
"learning_rate": 4.678296760308474e-05,
|
| 835 |
+
"loss": 120.44024658203125,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5929120628922259,
|
| 840 |
+
"grad_norm": 60.93242263793945,
|
| 841 |
+
"learning_rate": 4.670182560476948e-05,
|
| 842 |
+
"loss": 120.22978515625,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.5979367413913126,
|
| 847 |
+
"grad_norm": 59.64492416381836,
|
| 848 |
+
"learning_rate": 4.661974519552283e-05,
|
| 849 |
+
"loss": 108.0113525390625,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6029614198903992,
|
| 854 |
+
"grad_norm": 65.3152847290039,
|
| 855 |
+
"learning_rate": 4.653672992459242e-05,
|
| 856 |
+
"loss": 112.077978515625,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6079860983894858,
|
| 861 |
+
"grad_norm": 58.9239616394043,
|
| 862 |
+
"learning_rate": 4.645278338165035e-05,
|
| 863 |
+
"loss": 106.50635986328125,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.6130107768885725,
|
| 868 |
+
"grad_norm": 56.77158737182617,
|
| 869 |
+
"learning_rate": 4.636790919663794e-05,
|
| 870 |
+
"loss": 102.1367431640625,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6180354553876591,
|
| 875 |
+
"grad_norm": 49.77719497680664,
|
| 876 |
+
"learning_rate": 4.6282111039608784e-05,
|
| 877 |
+
"loss": 97.8525634765625,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.6230601338867459,
|
| 882 |
+
"grad_norm": 51.9918212890625,
|
| 883 |
+
"learning_rate": 4.6195392620570035e-05,
|
| 884 |
+
"loss": 105.98614501953125,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.6280848123858325,
|
| 889 |
+
"grad_norm": 58.79019546508789,
|
| 890 |
+
"learning_rate": 4.6107757689322014e-05,
|
| 891 |
+
"loss": 102.065966796875,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.6331094908849192,
|
| 896 |
+
"grad_norm": 49.530391693115234,
|
| 897 |
+
"learning_rate": 4.6019210035296034e-05,
|
| 898 |
+
"loss": 98.87720947265625,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.6381341693840058,
|
| 903 |
+
"grad_norm": 53.472782135009766,
|
| 904 |
+
"learning_rate": 4.592975348739054e-05,
|
| 905 |
+
"loss": 108.6241943359375,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.6431588478830925,
|
| 910 |
+
"grad_norm": 63.10646438598633,
|
| 911 |
+
"learning_rate": 4.584393142668341e-05,
|
| 912 |
+
"loss": 93.73190307617188,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6481835263821791,
|
| 917 |
+
"grad_norm": 59.84651565551758,
|
| 918 |
+
"learning_rate": 4.5752713697269135e-05,
|
| 919 |
+
"loss": 96.8425048828125,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.6532082048812659,
|
| 924 |
+
"grad_norm": 60.635169982910156,
|
| 925 |
+
"learning_rate": 4.5660598597571637e-05,
|
| 926 |
+
"loss": 103.2667236328125,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.6582328833803525,
|
| 931 |
+
"grad_norm": 61.401737213134766,
|
| 932 |
+
"learning_rate": 4.5567590110749646e-05,
|
| 933 |
+
"loss": 87.72919311523438,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.6632575618794391,
|
| 938 |
+
"grad_norm": 60.122772216796875,
|
| 939 |
+
"learning_rate": 4.547369225859297e-05,
|
| 940 |
+
"loss": 94.103662109375,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6682822403785258,
|
| 945 |
+
"grad_norm": 55.90855026245117,
|
| 946 |
+
"learning_rate": 4.537890910134854e-05,
|
| 947 |
+
"loss": 83.04338989257812,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.6733069188776124,
|
| 952 |
+
"grad_norm": 71.83133697509766,
|
| 953 |
+
"learning_rate": 4.528324473754486e-05,
|
| 954 |
+
"loss": 83.29258422851562,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.6783315973766991,
|
| 959 |
+
"grad_norm": 58.1335334777832,
|
| 960 |
+
"learning_rate": 4.518670330381479e-05,
|
| 961 |
+
"loss": 81.46260986328124,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.6833562758757857,
|
| 966 |
+
"grad_norm": 73.3138656616211,
|
| 967 |
+
"learning_rate": 4.5089288974716634e-05,
|
| 968 |
+
"loss": 81.26986694335938,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6883809543748725,
|
| 973 |
+
"grad_norm": 64.87886810302734,
|
| 974 |
+
"learning_rate": 4.49910059625537e-05,
|
| 975 |
+
"loss": 82.78674926757813,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.6934056328739591,
|
| 980 |
+
"grad_norm": 57.293861389160156,
|
| 981 |
+
"learning_rate": 4.489185851719208e-05,
|
| 982 |
+
"loss": 82.71456909179688,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.6984303113730458,
|
| 987 |
+
"grad_norm": 53.40017318725586,
|
| 988 |
+
"learning_rate": 4.4791850925876926e-05,
|
| 989 |
+
"loss": 84.79779052734375,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.7034549898721324,
|
| 994 |
+
"grad_norm": 48.285736083984375,
|
| 995 |
+
"learning_rate": 4.4690987513047044e-05,
|
| 996 |
+
"loss": 81.11317138671875,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.708479668371219,
|
| 1001 |
+
"grad_norm": 56.2747802734375,
|
| 1002 |
+
"learning_rate": 4.4589272640147915e-05,
|
| 1003 |
+
"loss": 72.64208374023437,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.7135043468703057,
|
| 1008 |
+
"grad_norm": 54.770931243896484,
|
| 1009 |
+
"learning_rate": 4.44867107054431e-05,
|
| 1010 |
+
"loss": 75.95457153320312,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.7185290253693923,
|
| 1015 |
+
"grad_norm": 49.85646057128906,
|
| 1016 |
+
"learning_rate": 4.4383306143824035e-05,
|
| 1017 |
+
"loss": 74.01212158203126,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.7235537038684791,
|
| 1022 |
+
"grad_norm": 49.286746978759766,
|
| 1023 |
+
"learning_rate": 4.427906342661828e-05,
|
| 1024 |
+
"loss": 67.97941284179687,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.7285783823675657,
|
| 1029 |
+
"grad_norm": 56.60735321044922,
|
| 1030 |
+
"learning_rate": 4.4173987061396186e-05,
|
| 1031 |
+
"loss": 70.84263916015625,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.7336030608666524,
|
| 1036 |
+
"grad_norm": 57.23542404174805,
|
| 1037 |
+
"learning_rate": 4.406808159177593e-05,
|
| 1038 |
+
"loss": 65.6290283203125,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.738627739365739,
|
| 1043 |
+
"grad_norm": 83.55549621582031,
|
| 1044 |
+
"learning_rate": 4.396135159722711e-05,
|
| 1045 |
+
"loss": 65.52483520507812,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.7436524178648257,
|
| 1050 |
+
"grad_norm": 53.960811614990234,
|
| 1051 |
+
"learning_rate": 4.3853801692872686e-05,
|
| 1052 |
+
"loss": 67.09555053710938,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.7486770963639123,
|
| 1057 |
+
"grad_norm": 45.474632263183594,
|
| 1058 |
+
"learning_rate": 4.374543652928943e-05,
|
| 1059 |
+
"loss": 67.3612548828125,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.753701774862999,
|
| 1064 |
+
"grad_norm": 56.11079025268555,
|
| 1065 |
+
"learning_rate": 4.363626079230681e-05,
|
| 1066 |
+
"loss": 62.279473876953126,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.7587264533620857,
|
| 1071 |
+
"grad_norm": 49.5059928894043,
|
| 1072 |
+
"learning_rate": 4.3526279202804387e-05,
|
| 1073 |
+
"loss": 73.84429321289062,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.7637511318611723,
|
| 1078 |
+
"grad_norm": 53.92646789550781,
|
| 1079 |
+
"learning_rate": 4.3415496516507695e-05,
|
| 1080 |
+
"loss": 61.59212646484375,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.768775810360259,
|
| 1085 |
+
"grad_norm": 57.92325210571289,
|
| 1086 |
+
"learning_rate": 4.3303917523782546e-05,
|
| 1087 |
+
"loss": 65.36521606445312,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.7738004888593456,
|
| 1092 |
+
"grad_norm": 47.250648498535156,
|
| 1093 |
+
"learning_rate": 4.3191547049427954e-05,
|
| 1094 |
+
"loss": 66.87857055664062,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7788251673584323,
|
| 1099 |
+
"grad_norm": 91.52741241455078,
|
| 1100 |
+
"learning_rate": 4.307838995246745e-05,
|
| 1101 |
+
"loss": 62.80021362304687,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.7838498458575189,
|
| 1106 |
+
"grad_norm": 46.733856201171875,
|
| 1107 |
+
"learning_rate": 4.296445112593902e-05,
|
| 1108 |
+
"loss": 55.04381103515625,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.7888745243566057,
|
| 1113 |
+
"grad_norm": 44.601497650146484,
|
| 1114 |
+
"learning_rate": 4.284973549668347e-05,
|
| 1115 |
+
"loss": 55.08817749023437,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.7938992028556923,
|
| 1120 |
+
"grad_norm": 50.94633102416992,
|
| 1121 |
+
"learning_rate": 4.273424802513145e-05,
|
| 1122 |
+
"loss": 52.93193359375,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.798923881354779,
|
| 1127 |
+
"grad_norm": 43.45872497558594,
|
| 1128 |
+
"learning_rate": 4.2617993705088886e-05,
|
| 1129 |
+
"loss": 54.57942504882813,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.8039485598538656,
|
| 1134 |
+
"grad_norm": 46.4177131652832,
|
| 1135 |
+
"learning_rate": 4.2500977563521115e-05,
|
| 1136 |
+
"loss": 54.19765014648438,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.8089732383529522,
|
| 1141 |
+
"grad_norm": 48.374786376953125,
|
| 1142 |
+
"learning_rate": 4.238320466033546e-05,
|
| 1143 |
+
"loss": 51.5730712890625,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.8139979168520389,
|
| 1148 |
+
"grad_norm": 51.431087493896484,
|
| 1149 |
+
"learning_rate": 4.2264680088162456e-05,
|
| 1150 |
+
"loss": 52.94944458007812,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.8190225953511255,
|
| 1155 |
+
"grad_norm": 40.70262145996094,
|
| 1156 |
+
"learning_rate": 4.214540897213565e-05,
|
| 1157 |
+
"loss": 53.7551513671875,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.8240472738502123,
|
| 1162 |
+
"grad_norm": 43.26173782348633,
|
| 1163 |
+
"learning_rate": 4.202539646966993e-05,
|
| 1164 |
+
"loss": 55.24525756835938,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.8290719523492989,
|
| 1169 |
+
"grad_norm": 47.32640838623047,
|
| 1170 |
+
"learning_rate": 4.190464777023861e-05,
|
| 1171 |
+
"loss": 48.86954040527344,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.8340966308483856,
|
| 1176 |
+
"grad_norm": 47.30034255981445,
|
| 1177 |
+
"learning_rate": 4.178316809514893e-05,
|
| 1178 |
+
"loss": 46.55088806152344,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.8391213093474722,
|
| 1183 |
+
"grad_norm": 51.20639419555664,
|
| 1184 |
+
"learning_rate": 4.1660962697316346e-05,
|
| 1185 |
+
"loss": 48.73444519042969,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
}
|
| 1188 |
+
],
|
| 1189 |
+
"logging_steps": 20,
|
| 1190 |
+
"max_steps": 9555,
|
| 1191 |
+
"num_input_tokens_seen": 0,
|
| 1192 |
+
"num_train_epochs": 3,
|
| 1193 |
+
"save_steps": 1000000000,
|
| 1194 |
+
"stateful_callbacks": {
|
| 1195 |
+
"TrainerControl": {
|
| 1196 |
+
"args": {
|
| 1197 |
+
"should_epoch_stop": false,
|
| 1198 |
+
"should_evaluate": false,
|
| 1199 |
+
"should_log": false,
|
| 1200 |
+
"should_save": true,
|
| 1201 |
+
"should_training_stop": false
|
| 1202 |
+
},
|
| 1203 |
+
"attributes": {}
|
| 1204 |
+
}
|
| 1205 |
+
},
|
| 1206 |
+
"total_flos": 7225991505137664.0,
|
| 1207 |
+
"train_batch_size": 1,
|
| 1208 |
+
"trial_name": null,
|
| 1209 |
+
"trial_params": null
|
| 1210 |
+
}
|