Upload folder using huggingface_hub
Browse files
slots/13/checkpoint-8362/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": 2,
|
| 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/13/checkpoint-8362/trainer_state.json
ADDED
|
@@ -0,0 +1,2967 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.6255728824204834,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 8362,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00041872211245305734,
|
| 14 |
+
"grad_norm": NaN,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 0.6013199090957642,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.008374442249061146,
|
| 21 |
+
"grad_norm": 1.9934037923812866,
|
| 22 |
+
"learning_rate": 4.999949009907666e-05,
|
| 23 |
+
"loss": 1.2950001766807155,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.016748884498122292,
|
| 28 |
+
"grad_norm": 1.5410668849945068,
|
| 29 |
+
"learning_rate": 4.999758461765953e-05,
|
| 30 |
+
"loss": 0.9534212112426758,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.025123326747183438,
|
| 35 |
+
"grad_norm": 1.7543447017669678,
|
| 36 |
+
"learning_rate": 4.999446714274127e-05,
|
| 37 |
+
"loss": 1.7281543731689453,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.033497768996244584,
|
| 42 |
+
"grad_norm": 1.228683590888977,
|
| 43 |
+
"learning_rate": 4.999007608070043e-05,
|
| 44 |
+
"loss": 1.0517969131469727,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.041872211245305734,
|
| 49 |
+
"grad_norm": 1.8294340372085571,
|
| 50 |
+
"learning_rate": 4.9984078259584456e-05,
|
| 51 |
+
"loss": 0.6183329105377198,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.050246653494366876,
|
| 56 |
+
"grad_norm": 2.365013837814331,
|
| 57 |
+
"learning_rate": 4.997666984880032e-05,
|
| 58 |
+
"loss": 0.6331160545349122,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.058621095743428026,
|
| 63 |
+
"grad_norm": 1.479344367980957,
|
| 64 |
+
"learning_rate": 4.9968325679939295e-05,
|
| 65 |
+
"loss": 0.7190958023071289,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.06699553799248917,
|
| 70 |
+
"grad_norm": 1.2224725484848022,
|
| 71 |
+
"learning_rate": 4.995816789488419e-05,
|
| 72 |
+
"loss": 0.9246182441711426,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07536998024155032,
|
| 77 |
+
"grad_norm": 1.1624501943588257,
|
| 78 |
+
"learning_rate": 4.994660098304831e-05,
|
| 79 |
+
"loss": 0.801685905456543,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.08374442249061147,
|
| 84 |
+
"grad_norm": 1.4690343141555786,
|
| 85 |
+
"learning_rate": 4.993362559749423e-05,
|
| 86 |
+
"loss": 0.6664809226989746,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09211886473967261,
|
| 91 |
+
"grad_norm": 1.4142041206359863,
|
| 92 |
+
"learning_rate": 4.991999504877833e-05,
|
| 93 |
+
"loss": 0.9447656631469726,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10049330698873375,
|
| 98 |
+
"grad_norm": 1.426574468612671,
|
| 99 |
+
"learning_rate": 4.9904275318961756e-05,
|
| 100 |
+
"loss": 0.5283678054809571,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.10886774923779491,
|
| 105 |
+
"grad_norm": 1.2339314222335815,
|
| 106 |
+
"learning_rate": 4.988803917569756e-05,
|
| 107 |
+
"loss": 0.9661962509155273,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.11724219148685605,
|
| 112 |
+
"grad_norm": 1.0341933965682983,
|
| 113 |
+
"learning_rate": 4.986957847640217e-05,
|
| 114 |
+
"loss": 0.5136320114135742,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1256166337359172,
|
| 119 |
+
"grad_norm": 1.5590699911117554,
|
| 120 |
+
"learning_rate": 4.984971365204418e-05,
|
| 121 |
+
"loss": 0.566891860961914,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13399107598497834,
|
| 126 |
+
"grad_norm": 2.064842700958252,
|
| 127 |
+
"learning_rate": 4.982844582418252e-05,
|
| 128 |
+
"loss": 0.6304716110229492,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14236551823403948,
|
| 133 |
+
"grad_norm": 1.4325364828109741,
|
| 134 |
+
"learning_rate": 4.980577619358906e-05,
|
| 135 |
+
"loss": 0.7040385723114013,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.15073996048310065,
|
| 140 |
+
"grad_norm": 1.0557610988616943,
|
| 141 |
+
"learning_rate": 4.978170604018079e-05,
|
| 142 |
+
"loss": 0.6057440757751464,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1591144027321618,
|
| 147 |
+
"grad_norm": 1.2590022087097168,
|
| 148 |
+
"learning_rate": 4.9756236722947616e-05,
|
| 149 |
+
"loss": 0.4318533420562744,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.16748884498122293,
|
| 154 |
+
"grad_norm": 1.700644850730896,
|
| 155 |
+
"learning_rate": 4.972936967987556e-05,
|
| 156 |
+
"loss": 0.8144928932189941,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.17586328723028408,
|
| 161 |
+
"grad_norm": 0.6783745884895325,
|
| 162 |
+
"learning_rate": 4.970110642786564e-05,
|
| 163 |
+
"loss": 0.5173389911651611,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.18423772947934522,
|
| 168 |
+
"grad_norm": 0.6910356879234314,
|
| 169 |
+
"learning_rate": 4.967144856264818e-05,
|
| 170 |
+
"loss": 0.4949141502380371,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.19261217172840636,
|
| 175 |
+
"grad_norm": 0.8224261999130249,
|
| 176 |
+
"learning_rate": 4.9640397758692715e-05,
|
| 177 |
+
"loss": 0.4392205238342285,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.2009866139774675,
|
| 182 |
+
"grad_norm": 1.6215633153915405,
|
| 183 |
+
"learning_rate": 4.960795576911349e-05,
|
| 184 |
+
"loss": 0.4062031745910645,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.20936105622652867,
|
| 189 |
+
"grad_norm": 0.8354963660240173,
|
| 190 |
+
"learning_rate": 4.9574124425570436e-05,
|
| 191 |
+
"loss": 0.34378366470336913,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.21773549847558982,
|
| 196 |
+
"grad_norm": 1.0993494987487793,
|
| 197 |
+
"learning_rate": 4.9538905638165805e-05,
|
| 198 |
+
"loss": 0.5852083206176758,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.22610994072465096,
|
| 203 |
+
"grad_norm": 2.658684015274048,
|
| 204 |
+
"learning_rate": 4.950230139533626e-05,
|
| 205 |
+
"loss": 0.4075888156890869,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.2344843829737121,
|
| 210 |
+
"grad_norm": 0.9528117179870605,
|
| 211 |
+
"learning_rate": 4.946431376374067e-05,
|
| 212 |
+
"loss": 0.3786322832107544,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.24285882522277324,
|
| 217 |
+
"grad_norm": 2.995469093322754,
|
| 218 |
+
"learning_rate": 4.9424944888143395e-05,
|
| 219 |
+
"loss": 0.5390625,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.2512332674718344,
|
| 224 |
+
"grad_norm": 0.8218050599098206,
|
| 225 |
+
"learning_rate": 4.9384196991293205e-05,
|
| 226 |
+
"loss": 0.39401955604553224,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.25960770972089553,
|
| 231 |
+
"grad_norm": 1.5551509857177734,
|
| 232 |
+
"learning_rate": 4.9342072373797774e-05,
|
| 233 |
+
"loss": 0.47962727546691897,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2679821519699567,
|
| 238 |
+
"grad_norm": 1.12935209274292,
|
| 239 |
+
"learning_rate": 4.9298573413993806e-05,
|
| 240 |
+
"loss": 0.42098431587219237,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2763565942190178,
|
| 245 |
+
"grad_norm": 2.1346328258514404,
|
| 246 |
+
"learning_rate": 4.925370256781273e-05,
|
| 247 |
+
"loss": 0.4414808750152588,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.28473103646807896,
|
| 252 |
+
"grad_norm": 1.285367727279663,
|
| 253 |
+
"learning_rate": 4.920746236864208e-05,
|
| 254 |
+
"loss": 0.4640543460845947,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.29310547871714016,
|
| 259 |
+
"grad_norm": 0.7970554232597351,
|
| 260 |
+
"learning_rate": 4.915985542718241e-05,
|
| 261 |
+
"loss": 0.3569347381591797,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3014799209662013,
|
| 266 |
+
"grad_norm": 0.6530137658119202,
|
| 267 |
+
"learning_rate": 4.9110884431299935e-05,
|
| 268 |
+
"loss": 0.41652464866638184,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.30985436321526244,
|
| 273 |
+
"grad_norm": 0.956561267375946,
|
| 274 |
+
"learning_rate": 4.906055214587476e-05,
|
| 275 |
+
"loss": 0.3489442110061646,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3182288054643236,
|
| 280 |
+
"grad_norm": 1.9765814542770386,
|
| 281 |
+
"learning_rate": 4.9008861412644756e-05,
|
| 282 |
+
"loss": 0.40424065589904784,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3266032477133847,
|
| 287 |
+
"grad_norm": 1.658229112625122,
|
| 288 |
+
"learning_rate": 4.895581515004517e-05,
|
| 289 |
+
"loss": 0.31643118858337405,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.33497768996244587,
|
| 294 |
+
"grad_norm": 2.2325282096862793,
|
| 295 |
+
"learning_rate": 4.89014163530438e-05,
|
| 296 |
+
"loss": 0.3643105745315552,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.343352132211507,
|
| 301 |
+
"grad_norm": 0.4895799160003662,
|
| 302 |
+
"learning_rate": 4.8845668092971934e-05,
|
| 303 |
+
"loss": 0.3214404821395874,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.35172657446056815,
|
| 308 |
+
"grad_norm": 1.4924691915512085,
|
| 309 |
+
"learning_rate": 4.87885735173509e-05,
|
| 310 |
+
"loss": 0.4828330993652344,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3601010167096293,
|
| 315 |
+
"grad_norm": 0.6766088604927063,
|
| 316 |
+
"learning_rate": 4.873013584971442e-05,
|
| 317 |
+
"loss": 0.3910247325897217,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.36847545895869044,
|
| 322 |
+
"grad_norm": 0.48699629306793213,
|
| 323 |
+
"learning_rate": 4.867035838942657e-05,
|
| 324 |
+
"loss": 0.2918583631515503,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3768499012077516,
|
| 329 |
+
"grad_norm": 1.2841662168502808,
|
| 330 |
+
"learning_rate": 4.86092445114955e-05,
|
| 331 |
+
"loss": 0.2856093406677246,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3852243434568127,
|
| 336 |
+
"grad_norm": 0.21915394067764282,
|
| 337 |
+
"learning_rate": 4.8546797666382896e-05,
|
| 338 |
+
"loss": 0.33952271938323975,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.39359878570587387,
|
| 343 |
+
"grad_norm": 0.6552733778953552,
|
| 344 |
+
"learning_rate": 4.8483021379809154e-05,
|
| 345 |
+
"loss": 0.3813821315765381,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.401973227954935,
|
| 350 |
+
"grad_norm": 0.4852622449398041,
|
| 351 |
+
"learning_rate": 4.8417919252554354e-05,
|
| 352 |
+
"loss": 0.3341342926025391,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4103476702039962,
|
| 357 |
+
"grad_norm": 1.5116126537322998,
|
| 358 |
+
"learning_rate": 4.8351494960254916e-05,
|
| 359 |
+
"loss": 0.29020323753356936,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.41872211245305735,
|
| 364 |
+
"grad_norm": 0.4594159424304962,
|
| 365 |
+
"learning_rate": 4.8283752253196094e-05,
|
| 366 |
+
"loss": 0.25537676811218263,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4270965547021185,
|
| 371 |
+
"grad_norm": 0.8960456252098083,
|
| 372 |
+
"learning_rate": 4.8214694956100245e-05,
|
| 373 |
+
"loss": 0.26440722942352296,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.43547099695117963,
|
| 378 |
+
"grad_norm": 0.6404821872711182,
|
| 379 |
+
"learning_rate": 4.814432696791089e-05,
|
| 380 |
+
"loss": 0.2976436376571655,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4438454392002408,
|
| 385 |
+
"grad_norm": 1.3145599365234375,
|
| 386 |
+
"learning_rate": 4.807265226157255e-05,
|
| 387 |
+
"loss": 0.5187122344970703,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.4522198814493019,
|
| 392 |
+
"grad_norm": 0.7199428081512451,
|
| 393 |
+
"learning_rate": 4.7999674883806465e-05,
|
| 394 |
+
"loss": 0.3027026653289795,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.46059432369836306,
|
| 399 |
+
"grad_norm": 0.2537328600883484,
|
| 400 |
+
"learning_rate": 4.792539895488213e-05,
|
| 401 |
+
"loss": 0.44816179275512696,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4689687659474242,
|
| 406 |
+
"grad_norm": 0.6229515075683594,
|
| 407 |
+
"learning_rate": 4.784982866838462e-05,
|
| 408 |
+
"loss": 0.2838144063949585,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.47734320819648535,
|
| 413 |
+
"grad_norm": 0.32296815514564514,
|
| 414 |
+
"learning_rate": 4.777296829097785e-05,
|
| 415 |
+
"loss": 0.4447695255279541,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.4857176504455465,
|
| 420 |
+
"grad_norm": 0.8773243427276611,
|
| 421 |
+
"learning_rate": 4.769482216216369e-05,
|
| 422 |
+
"loss": 0.3049264192581177,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.49409209269460763,
|
| 427 |
+
"grad_norm": 6.263951301574707,
|
| 428 |
+
"learning_rate": 4.761539469403695e-05,
|
| 429 |
+
"loss": 0.2543496131896973,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5024665349436688,
|
| 434 |
+
"grad_norm": 0.31666627526283264,
|
| 435 |
+
"learning_rate": 4.7534690371036274e-05,
|
| 436 |
+
"loss": 0.22089865207672119,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5108409771927299,
|
| 441 |
+
"grad_norm": 1.305454134941101,
|
| 442 |
+
"learning_rate": 4.7452713749690934e-05,
|
| 443 |
+
"loss": 0.30597648620605467,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5192154194417911,
|
| 448 |
+
"grad_norm": 0.7708957195281982,
|
| 449 |
+
"learning_rate": 4.73694694583636e-05,
|
| 450 |
+
"loss": 0.31495089530944825,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5275898616908522,
|
| 455 |
+
"grad_norm": 1.2214106321334839,
|
| 456 |
+
"learning_rate": 4.728496219698902e-05,
|
| 457 |
+
"loss": 0.2793067455291748,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5359643039399133,
|
| 462 |
+
"grad_norm": 0.6414151191711426,
|
| 463 |
+
"learning_rate": 4.719919673680863e-05,
|
| 464 |
+
"loss": 0.24640004634857177,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5443387461889745,
|
| 469 |
+
"grad_norm": 0.4211215376853943,
|
| 470 |
+
"learning_rate": 4.711217792010122e-05,
|
| 471 |
+
"loss": 0.21954472064971925,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5527131884380356,
|
| 476 |
+
"grad_norm": 1.1475203037261963,
|
| 477 |
+
"learning_rate": 4.70239106599095e-05,
|
| 478 |
+
"loss": 0.29560739994049073,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5610876306870968,
|
| 483 |
+
"grad_norm": 1.3786718845367432,
|
| 484 |
+
"learning_rate": 4.6934399939762746e-05,
|
| 485 |
+
"loss": 0.2873764276504517,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5694620729361579,
|
| 490 |
+
"grad_norm": 1.0504395961761475,
|
| 491 |
+
"learning_rate": 4.684365081339541e-05,
|
| 492 |
+
"loss": 0.35107598304748533,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5778365151852191,
|
| 497 |
+
"grad_norm": 2.374439001083374,
|
| 498 |
+
"learning_rate": 4.675166840446179e-05,
|
| 499 |
+
"loss": 0.33103017807006835,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5862109574342803,
|
| 504 |
+
"grad_norm": 0.30569833517074585,
|
| 505 |
+
"learning_rate": 4.6658457906246754e-05,
|
| 506 |
+
"loss": 0.15970700979232788,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5945853996833415,
|
| 511 |
+
"grad_norm": 2.7184505462646484,
|
| 512 |
+
"learning_rate": 4.656402458137255e-05,
|
| 513 |
+
"loss": 0.18424086570739745,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6029598419324026,
|
| 518 |
+
"grad_norm": 0.9072023630142212,
|
| 519 |
+
"learning_rate": 4.646837376150164e-05,
|
| 520 |
+
"loss": 0.2588572263717651,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6113342841814637,
|
| 525 |
+
"grad_norm": 0.4133734405040741,
|
| 526 |
+
"learning_rate": 4.637151084703571e-05,
|
| 527 |
+
"loss": 0.20942835807800292,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6197087264305249,
|
| 532 |
+
"grad_norm": 0.8726677298545837,
|
| 533 |
+
"learning_rate": 4.6273441306810724e-05,
|
| 534 |
+
"loss": 0.36864147186279295,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.628083168679586,
|
| 539 |
+
"grad_norm": 0.9570859670639038,
|
| 540 |
+
"learning_rate": 4.617417067778825e-05,
|
| 541 |
+
"loss": 0.35473387241363524,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6364576109286472,
|
| 546 |
+
"grad_norm": 1.6990162134170532,
|
| 547 |
+
"learning_rate": 4.6073704564742734e-05,
|
| 548 |
+
"loss": 0.28413536548614504,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6448320531777083,
|
| 553 |
+
"grad_norm": 0.5367034673690796,
|
| 554 |
+
"learning_rate": 4.5972048639945105e-05,
|
| 555 |
+
"loss": 0.23524901866912842,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6532064954267695,
|
| 560 |
+
"grad_norm": 0.1774810403585434,
|
| 561 |
+
"learning_rate": 4.5869208642842536e-05,
|
| 562 |
+
"loss": 0.16743130683898927,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.6615809376758306,
|
| 567 |
+
"grad_norm": 0.4111934304237366,
|
| 568 |
+
"learning_rate": 4.5765190379734384e-05,
|
| 569 |
+
"loss": 0.2366396427154541,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.6699553799248917,
|
| 574 |
+
"grad_norm": 1.3504371643066406,
|
| 575 |
+
"learning_rate": 4.565999972344436e-05,
|
| 576 |
+
"loss": 0.17410001754760743,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6783298221739529,
|
| 581 |
+
"grad_norm": 0.26619571447372437,
|
| 582 |
+
"learning_rate": 4.5553642612988954e-05,
|
| 583 |
+
"loss": 0.21413214206695558,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.686704264423014,
|
| 588 |
+
"grad_norm": 1.2675085067749023,
|
| 589 |
+
"learning_rate": 4.544612505324213e-05,
|
| 590 |
+
"loss": 0.22462894916534423,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6950787066720752,
|
| 595 |
+
"grad_norm": 0.2958528995513916,
|
| 596 |
+
"learning_rate": 4.533745311459631e-05,
|
| 597 |
+
"loss": 0.17030240297317506,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7034531489211363,
|
| 602 |
+
"grad_norm": 0.29442816972732544,
|
| 603 |
+
"learning_rate": 4.5227632932619604e-05,
|
| 604 |
+
"loss": 0.30537502765655516,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7118275911701974,
|
| 609 |
+
"grad_norm": 1.2775323390960693,
|
| 610 |
+
"learning_rate": 4.5116670707709396e-05,
|
| 611 |
+
"loss": 0.21231284141540527,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7202020334192586,
|
| 616 |
+
"grad_norm": 0.6058893203735352,
|
| 617 |
+
"learning_rate": 4.5004572704742345e-05,
|
| 618 |
+
"loss": 0.2254181146621704,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7285764756683197,
|
| 623 |
+
"grad_norm": 0.67788165807724,
|
| 624 |
+
"learning_rate": 4.489134525272057e-05,
|
| 625 |
+
"loss": 0.2267812728881836,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7369509179173809,
|
| 630 |
+
"grad_norm": 0.5695115923881531,
|
| 631 |
+
"learning_rate": 4.477699474441441e-05,
|
| 632 |
+
"loss": 0.2287515878677368,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.745325360166442,
|
| 637 |
+
"grad_norm": 1.9975258111953735,
|
| 638 |
+
"learning_rate": 4.466152763600142e-05,
|
| 639 |
+
"loss": 0.15261965990066528,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7536998024155032,
|
| 644 |
+
"grad_norm": 0.8055058717727661,
|
| 645 |
+
"learning_rate": 4.454495044670189e-05,
|
| 646 |
+
"loss": 0.2953073740005493,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7620742446645643,
|
| 651 |
+
"grad_norm": 0.3822990357875824,
|
| 652 |
+
"learning_rate": 4.4427269758410806e-05,
|
| 653 |
+
"loss": 0.22958216667175294,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7704486869136254,
|
| 658 |
+
"grad_norm": 1.03629469871521,
|
| 659 |
+
"learning_rate": 4.4308492215326145e-05,
|
| 660 |
+
"loss": 0.2106247663497925,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.7788231291626866,
|
| 665 |
+
"grad_norm": 0.15963371098041534,
|
| 666 |
+
"learning_rate": 4.418862452357387e-05,
|
| 667 |
+
"loss": 0.14056196212768554,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.7871975714117477,
|
| 672 |
+
"grad_norm": 0.46281635761260986,
|
| 673 |
+
"learning_rate": 4.4067673450829184e-05,
|
| 674 |
+
"loss": 0.2616506814956665,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.7955720136608089,
|
| 679 |
+
"grad_norm": 0.8718962669372559,
|
| 680 |
+
"learning_rate": 4.394564582593452e-05,
|
| 681 |
+
"loss": 0.20136590003967286,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.80394645590987,
|
| 686 |
+
"grad_norm": 0.16621536016464233,
|
| 687 |
+
"learning_rate": 4.382254853851396e-05,
|
| 688 |
+
"loss": 0.1678437829017639,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8123208981589312,
|
| 693 |
+
"grad_norm": 0.31210511922836304,
|
| 694 |
+
"learning_rate": 4.3698388538584214e-05,
|
| 695 |
+
"loss": 0.1877954602241516,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8206953404079924,
|
| 700 |
+
"grad_norm": 0.22836421430110931,
|
| 701 |
+
"learning_rate": 4.357317283616228e-05,
|
| 702 |
+
"loss": 0.1965322971343994,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8290697826570536,
|
| 707 |
+
"grad_norm": 1.5269449949264526,
|
| 708 |
+
"learning_rate": 4.3446908500869624e-05,
|
| 709 |
+
"loss": 0.14575521945953368,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8374442249061147,
|
| 714 |
+
"grad_norm": 0.541411817073822,
|
| 715 |
+
"learning_rate": 4.3319602661533054e-05,
|
| 716 |
+
"loss": 0.1215706706047058,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8458186671551758,
|
| 721 |
+
"grad_norm": 1.0127756595611572,
|
| 722 |
+
"learning_rate": 4.319126250578223e-05,
|
| 723 |
+
"loss": 0.1414382576942444,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.854193109404237,
|
| 728 |
+
"grad_norm": 0.496611624956131,
|
| 729 |
+
"learning_rate": 4.306189527964382e-05,
|
| 730 |
+
"loss": 0.1537245512008667,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8625675516532981,
|
| 735 |
+
"grad_norm": 0.17574390769004822,
|
| 736 |
+
"learning_rate": 4.2931508287132435e-05,
|
| 737 |
+
"loss": 0.14648995399475098,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.8709419939023593,
|
| 742 |
+
"grad_norm": 1.1143279075622559,
|
| 743 |
+
"learning_rate": 4.2800108889838244e-05,
|
| 744 |
+
"loss": 0.1767507791519165,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8793164361514204,
|
| 749 |
+
"grad_norm": 0.40195536613464355,
|
| 750 |
+
"learning_rate": 4.26677045065113e-05,
|
| 751 |
+
"loss": 0.12700635194778442,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.8876908784004816,
|
| 756 |
+
"grad_norm": 0.5011544823646545,
|
| 757 |
+
"learning_rate": 4.253430261264272e-05,
|
| 758 |
+
"loss": 0.1776489496231079,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8960653206495427,
|
| 763 |
+
"grad_norm": 0.9425447583198547,
|
| 764 |
+
"learning_rate": 4.239991074004262e-05,
|
| 765 |
+
"loss": 0.17866064310073854,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9044397628986038,
|
| 770 |
+
"grad_norm": 1.5652395486831665,
|
| 771 |
+
"learning_rate": 4.226453647641484e-05,
|
| 772 |
+
"loss": 0.18456958532333373,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.912814205147665,
|
| 777 |
+
"grad_norm": 0.9656435251235962,
|
| 778 |
+
"learning_rate": 4.2128187464928604e-05,
|
| 779 |
+
"loss": 0.20541505813598632,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9211886473967261,
|
| 784 |
+
"grad_norm": 0.20453469455242157,
|
| 785 |
+
"learning_rate": 4.1990871403786934e-05,
|
| 786 |
+
"loss": 0.14855676889419556,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9295630896457873,
|
| 791 |
+
"grad_norm": 0.19641469419002533,
|
| 792 |
+
"learning_rate": 4.185259604579203e-05,
|
| 793 |
+
"loss": 0.09722235202789306,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.9379375318948484,
|
| 798 |
+
"grad_norm": 0.34978342056274414,
|
| 799 |
+
"learning_rate": 4.171336919790759e-05,
|
| 800 |
+
"loss": 0.1498737931251526,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9463119741439096,
|
| 805 |
+
"grad_norm": 1.2353084087371826,
|
| 806 |
+
"learning_rate": 4.1573198720817955e-05,
|
| 807 |
+
"loss": 0.17144311666488649,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9546864163929707,
|
| 812 |
+
"grad_norm": 1.176924228668213,
|
| 813 |
+
"learning_rate": 4.143209252848433e-05,
|
| 814 |
+
"loss": 0.20422227382659913,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9630608586420318,
|
| 819 |
+
"grad_norm": 0.27958446741104126,
|
| 820 |
+
"learning_rate": 4.1290058587698054e-05,
|
| 821 |
+
"loss": 0.1507509708404541,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.971435300891093,
|
| 826 |
+
"grad_norm": 1.0945500135421753,
|
| 827 |
+
"learning_rate": 4.114710491763063e-05,
|
| 828 |
+
"loss": 0.21894502639770508,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9798097431401541,
|
| 833 |
+
"grad_norm": 0.3587752878665924,
|
| 834 |
+
"learning_rate": 4.100323958938112e-05,
|
| 835 |
+
"loss": 0.09666566252708435,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9881841853892153,
|
| 840 |
+
"grad_norm": 1.3135223388671875,
|
| 841 |
+
"learning_rate": 4.085847072552036e-05,
|
| 842 |
+
"loss": 0.16662789583206178,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.9965586276382764,
|
| 847 |
+
"grad_norm": 0.44114890694618225,
|
| 848 |
+
"learning_rate": 4.0712806499632397e-05,
|
| 849 |
+
"loss": 0.1317327857017517,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.0046059432369836,
|
| 854 |
+
"grad_norm": 0.4595654308795929,
|
| 855 |
+
"learning_rate": 4.0566255135853014e-05,
|
| 856 |
+
"loss": 0.16484848260879517,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.0129803854860449,
|
| 861 |
+
"grad_norm": 0.2524283528327942,
|
| 862 |
+
"learning_rate": 4.04188249084054e-05,
|
| 863 |
+
"loss": 0.09629667997360229,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.0213548277351059,
|
| 868 |
+
"grad_norm": 0.640849769115448,
|
| 869 |
+
"learning_rate": 4.0270524141133e-05,
|
| 870 |
+
"loss": 0.08449938893318176,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.0297292699841671,
|
| 875 |
+
"grad_norm": 0.9351454377174377,
|
| 876 |
+
"learning_rate": 4.01213612070295e-05,
|
| 877 |
+
"loss": 0.13373035192489624,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0381037122332282,
|
| 882 |
+
"grad_norm": 0.1494932770729065,
|
| 883 |
+
"learning_rate": 3.997134452776617e-05,
|
| 884 |
+
"loss": 0.09310936331748962,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0464781544822894,
|
| 889 |
+
"grad_norm": 0.7994132041931152,
|
| 890 |
+
"learning_rate": 3.982048257321631e-05,
|
| 891 |
+
"loss": 0.11848243474960327,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.0548525967313505,
|
| 896 |
+
"grad_norm": 1.789854884147644,
|
| 897 |
+
"learning_rate": 3.966878386097711e-05,
|
| 898 |
+
"loss": 0.08053499460220337,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.0632270389804117,
|
| 903 |
+
"grad_norm": 0.09759683907032013,
|
| 904 |
+
"learning_rate": 3.95162569558887e-05,
|
| 905 |
+
"loss": 0.13048415184020995,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0716014812294727,
|
| 910 |
+
"grad_norm": 1.745640754699707,
|
| 911 |
+
"learning_rate": 3.936291046955061e-05,
|
| 912 |
+
"loss": 0.06222975254058838,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.079975923478534,
|
| 917 |
+
"grad_norm": 0.9128367304801941,
|
| 918 |
+
"learning_rate": 3.920875305983553e-05,
|
| 919 |
+
"loss": 0.07491644620895385,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.088350365727595,
|
| 924 |
+
"grad_norm": 1.008280873298645,
|
| 925 |
+
"learning_rate": 3.905379343040054e-05,
|
| 926 |
+
"loss": 0.14277656078338624,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.0967248079766563,
|
| 931 |
+
"grad_norm": 0.48391154408454895,
|
| 932 |
+
"learning_rate": 3.889804033019567e-05,
|
| 933 |
+
"loss": 0.0745099127292633,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.1050992502257173,
|
| 938 |
+
"grad_norm": 0.2227552831172943,
|
| 939 |
+
"learning_rate": 3.8741502552969956e-05,
|
| 940 |
+
"loss": 0.09417952299118042,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.1134736924747786,
|
| 945 |
+
"grad_norm": 0.16306816041469574,
|
| 946 |
+
"learning_rate": 3.8584188936774925e-05,
|
| 947 |
+
"loss": 0.12839992046356202,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.1218481347238396,
|
| 952 |
+
"grad_norm": 0.07621652632951736,
|
| 953 |
+
"learning_rate": 3.842610836346563e-05,
|
| 954 |
+
"loss": 0.10739836692810059,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.1302225769729008,
|
| 959 |
+
"grad_norm": 0.5850180983543396,
|
| 960 |
+
"learning_rate": 3.8267269758199175e-05,
|
| 961 |
+
"loss": 0.07980061173439026,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.1385970192219619,
|
| 966 |
+
"grad_norm": 2.3212532997131348,
|
| 967 |
+
"learning_rate": 3.8107682088930794e-05,
|
| 968 |
+
"loss": 0.08940492272377014,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.1469714614710231,
|
| 973 |
+
"grad_norm": 0.10123835504055023,
|
| 974 |
+
"learning_rate": 3.7947354365907536e-05,
|
| 975 |
+
"loss": 0.09007878303527832,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.1553459037200842,
|
| 980 |
+
"grad_norm": 0.9358357191085815,
|
| 981 |
+
"learning_rate": 3.7786295641159544e-05,
|
| 982 |
+
"loss": 0.08590203523635864,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.1637203459691454,
|
| 987 |
+
"grad_norm": 0.22766231000423431,
|
| 988 |
+
"learning_rate": 3.7624515007989004e-05,
|
| 989 |
+
"loss": 0.12319990396499633,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1720947882182067,
|
| 994 |
+
"grad_norm": 1.0723813772201538,
|
| 995 |
+
"learning_rate": 3.746202160045669e-05,
|
| 996 |
+
"loss": 0.07759063839912414,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.1804692304672677,
|
| 1001 |
+
"grad_norm": 0.2642482817173004,
|
| 1002 |
+
"learning_rate": 3.729882459286632e-05,
|
| 1003 |
+
"loss": 0.07289831042289734,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.1888436727163287,
|
| 1008 |
+
"grad_norm": 0.11382978409528732,
|
| 1009 |
+
"learning_rate": 3.713493319924655e-05,
|
| 1010 |
+
"loss": 0.08417699337005616,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.19721811496539,
|
| 1015 |
+
"grad_norm": 0.6055602431297302,
|
| 1016 |
+
"learning_rate": 3.697035667283072e-05,
|
| 1017 |
+
"loss": 0.0856675386428833,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.2055925572144512,
|
| 1022 |
+
"grad_norm": 0.12562307715415955,
|
| 1023 |
+
"learning_rate": 3.68051043055345e-05,
|
| 1024 |
+
"loss": 0.10780172348022461,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.2139669994635123,
|
| 1029 |
+
"grad_norm": 0.06781217455863953,
|
| 1030 |
+
"learning_rate": 3.66391854274312e-05,
|
| 1031 |
+
"loss": 0.055737143754959105,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.2223414417125735,
|
| 1036 |
+
"grad_norm": 0.23094801604747772,
|
| 1037 |
+
"learning_rate": 3.647260940622504e-05,
|
| 1038 |
+
"loss": 0.08827725648880005,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.2307158839616346,
|
| 1043 |
+
"grad_norm": 0.26578766107559204,
|
| 1044 |
+
"learning_rate": 3.6305385646722224e-05,
|
| 1045 |
+
"loss": 0.10403858423233033,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.2390903262106958,
|
| 1050 |
+
"grad_norm": 1.2067272663116455,
|
| 1051 |
+
"learning_rate": 3.613752359029998e-05,
|
| 1052 |
+
"loss": 0.09574756622314454,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.2474647684597568,
|
| 1057 |
+
"grad_norm": 0.28369495272636414,
|
| 1058 |
+
"learning_rate": 3.5969032714373495e-05,
|
| 1059 |
+
"loss": 0.12156195640563965,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.255839210708818,
|
| 1064 |
+
"grad_norm": 0.6278151273727417,
|
| 1065 |
+
"learning_rate": 3.57999225318608e-05,
|
| 1066 |
+
"loss": 0.09732474088668823,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.2642136529578791,
|
| 1071 |
+
"grad_norm": 0.7716904282569885,
|
| 1072 |
+
"learning_rate": 3.563020259064569e-05,
|
| 1073 |
+
"loss": 0.10135951042175292,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.2725880952069404,
|
| 1078 |
+
"grad_norm": 0.07069838047027588,
|
| 1079 |
+
"learning_rate": 3.5459882473038695e-05,
|
| 1080 |
+
"loss": 0.2272268295288086,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.2809625374560014,
|
| 1085 |
+
"grad_norm": 0.5034748911857605,
|
| 1086 |
+
"learning_rate": 3.528897179523598e-05,
|
| 1087 |
+
"loss": 0.07878611087799073,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.2893369797050627,
|
| 1092 |
+
"grad_norm": 0.1114577129483223,
|
| 1093 |
+
"learning_rate": 3.5117480206776506e-05,
|
| 1094 |
+
"loss": 0.07403420805931091,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.2977114219541237,
|
| 1099 |
+
"grad_norm": 0.1419895440340042,
|
| 1100 |
+
"learning_rate": 3.494541738999717e-05,
|
| 1101 |
+
"loss": 0.06619678139686584,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.306085864203185,
|
| 1106 |
+
"grad_norm": 0.3242378234863281,
|
| 1107 |
+
"learning_rate": 3.4772793059486176e-05,
|
| 1108 |
+
"loss": 0.07819657921791076,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.314460306452246,
|
| 1113 |
+
"grad_norm": 0.13734346628189087,
|
| 1114 |
+
"learning_rate": 3.4599616961534506e-05,
|
| 1115 |
+
"loss": 0.07510924339294434,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.3228347487013072,
|
| 1120 |
+
"grad_norm": 0.9328349232673645,
|
| 1121 |
+
"learning_rate": 3.44258988735857e-05,
|
| 1122 |
+
"loss": 0.07773483395576478,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.3312091909503683,
|
| 1127 |
+
"grad_norm": 0.14395828545093536,
|
| 1128 |
+
"learning_rate": 3.4251648603683824e-05,
|
| 1129 |
+
"loss": 0.07977863550186157,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.3395836331994295,
|
| 1134 |
+
"grad_norm": 0.4411904513835907,
|
| 1135 |
+
"learning_rate": 3.407687598991965e-05,
|
| 1136 |
+
"loss": 0.07483948469161987,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.3479580754484906,
|
| 1141 |
+
"grad_norm": 0.2287372499704361,
|
| 1142 |
+
"learning_rate": 3.390159089987529e-05,
|
| 1143 |
+
"loss": 0.08884957432746887,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.3563325176975518,
|
| 1148 |
+
"grad_norm": 0.7106556296348572,
|
| 1149 |
+
"learning_rate": 3.3725803230067006e-05,
|
| 1150 |
+
"loss": 0.10559600591659546,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3647069599466128,
|
| 1155 |
+
"grad_norm": 0.7979939579963684,
|
| 1156 |
+
"learning_rate": 3.354952290538649e-05,
|
| 1157 |
+
"loss": 0.11070560216903687,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.373081402195674,
|
| 1162 |
+
"grad_norm": 0.2569740414619446,
|
| 1163 |
+
"learning_rate": 3.337275987854052e-05,
|
| 1164 |
+
"loss": 0.0596321702003479,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.3814558444447353,
|
| 1169 |
+
"grad_norm": 0.16822633147239685,
|
| 1170 |
+
"learning_rate": 3.3195524129488986e-05,
|
| 1171 |
+
"loss": 0.08175230622291565,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.3898302866937964,
|
| 1176 |
+
"grad_norm": 0.1559642255306244,
|
| 1177 |
+
"learning_rate": 3.3017825664881496e-05,
|
| 1178 |
+
"loss": 0.07060136198997498,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.3982047289428574,
|
| 1183 |
+
"grad_norm": 0.565130352973938,
|
| 1184 |
+
"learning_rate": 3.283967451749233e-05,
|
| 1185 |
+
"loss": 0.06781386733055114,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.4065791711919187,
|
| 1190 |
+
"grad_norm": 0.05673038586974144,
|
| 1191 |
+
"learning_rate": 3.266108074565406e-05,
|
| 1192 |
+
"loss": 0.043686714768409726,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.41495361344098,
|
| 1197 |
+
"grad_norm": 0.5792385935783386,
|
| 1198 |
+
"learning_rate": 3.2482054432689616e-05,
|
| 1199 |
+
"loss": 0.05318898558616638,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.423328055690041,
|
| 1204 |
+
"grad_norm": 0.8359180688858032,
|
| 1205 |
+
"learning_rate": 3.2302605686343016e-05,
|
| 1206 |
+
"loss": 0.07649694681167603,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.431702497939102,
|
| 1211 |
+
"grad_norm": 0.11994684487581253,
|
| 1212 |
+
"learning_rate": 3.212274463820866e-05,
|
| 1213 |
+
"loss": 0.043499928712844846,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.4400769401881632,
|
| 1218 |
+
"grad_norm": 0.3480674624443054,
|
| 1219 |
+
"learning_rate": 3.194248144315933e-05,
|
| 1220 |
+
"loss": 0.07629216313362122,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.4484513824372245,
|
| 1225 |
+
"grad_norm": 0.349407821893692,
|
| 1226 |
+
"learning_rate": 3.176182627877284e-05,
|
| 1227 |
+
"loss": 0.06937903761863709,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.4568258246862855,
|
| 1232 |
+
"grad_norm": 0.32339081168174744,
|
| 1233 |
+
"learning_rate": 3.15807893447574e-05,
|
| 1234 |
+
"loss": 0.0664923369884491,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.4652002669353466,
|
| 1239 |
+
"grad_norm": 1.2784419059753418,
|
| 1240 |
+
"learning_rate": 3.1399380862375785e-05,
|
| 1241 |
+
"loss": 0.043991288542747496,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.4735747091844078,
|
| 1246 |
+
"grad_norm": 0.9719874262809753,
|
| 1247 |
+
"learning_rate": 3.121761107386817e-05,
|
| 1248 |
+
"loss": 0.0807355523109436,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.481949151433469,
|
| 1253 |
+
"grad_norm": 0.06417154520750046,
|
| 1254 |
+
"learning_rate": 3.103549024187398e-05,
|
| 1255 |
+
"loss": 0.06301149725914001,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.49032359368253,
|
| 1260 |
+
"grad_norm": 0.5472319722175598,
|
| 1261 |
+
"learning_rate": 3.085302864885235e-05,
|
| 1262 |
+
"loss": 0.13920137882232667,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4986980359315913,
|
| 1267 |
+
"grad_norm": 0.4602014720439911,
|
| 1268 |
+
"learning_rate": 3.067023659650163e-05,
|
| 1269 |
+
"loss": 0.10979901552200318,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.5070724781806524,
|
| 1274 |
+
"grad_norm": 0.48343533277511597,
|
| 1275 |
+
"learning_rate": 3.048712440517777e-05,
|
| 1276 |
+
"loss": 0.056929266452789305,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.5154469204297136,
|
| 1281 |
+
"grad_norm": 0.5795832276344299,
|
| 1282 |
+
"learning_rate": 3.030370241331159e-05,
|
| 1283 |
+
"loss": 0.06327124834060668,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.5238213626787747,
|
| 1288 |
+
"grad_norm": 0.07871934771537781,
|
| 1289 |
+
"learning_rate": 3.0119980976825156e-05,
|
| 1290 |
+
"loss": 0.05837821960449219,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.532195804927836,
|
| 1295 |
+
"grad_norm": 0.38906294107437134,
|
| 1296 |
+
"learning_rate": 2.9935970468546992e-05,
|
| 1297 |
+
"loss": 0.05847613215446472,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.540570247176897,
|
| 1302 |
+
"grad_norm": 1.1948798894882202,
|
| 1303 |
+
"learning_rate": 2.9751681277626507e-05,
|
| 1304 |
+
"loss": 0.06106034517288208,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.5489446894259582,
|
| 1309 |
+
"grad_norm": 0.06581711769104004,
|
| 1310 |
+
"learning_rate": 2.956712380894742e-05,
|
| 1311 |
+
"loss": 0.06192805767059326,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.5573191316750195,
|
| 1316 |
+
"grad_norm": 0.11205727607011795,
|
| 1317 |
+
"learning_rate": 2.9382308482540237e-05,
|
| 1318 |
+
"loss": 0.05574634075164795,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.5656935739240805,
|
| 1323 |
+
"grad_norm": 3.2353687286376953,
|
| 1324 |
+
"learning_rate": 2.919724573299406e-05,
|
| 1325 |
+
"loss": 0.07629644870758057,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.5740680161731415,
|
| 1330 |
+
"grad_norm": 0.07098077982664108,
|
| 1331 |
+
"learning_rate": 2.9011946008867342e-05,
|
| 1332 |
+
"loss": 0.030209916830062866,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.5824424584222028,
|
| 1337 |
+
"grad_norm": 0.28404733538627625,
|
| 1338 |
+
"learning_rate": 2.882641977209803e-05,
|
| 1339 |
+
"loss": 0.05649316906929016,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.590816900671264,
|
| 1344 |
+
"grad_norm": 0.46935683488845825,
|
| 1345 |
+
"learning_rate": 2.8640677497412877e-05,
|
| 1346 |
+
"loss": 0.05797313451766968,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.599191342920325,
|
| 1351 |
+
"grad_norm": 0.5720381736755371,
|
| 1352 |
+
"learning_rate": 2.8454729671736013e-05,
|
| 1353 |
+
"loss": 0.05965529084205627,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.607565785169386,
|
| 1358 |
+
"grad_norm": 0.1056889146566391,
|
| 1359 |
+
"learning_rate": 2.826858679359693e-05,
|
| 1360 |
+
"loss": 0.06270216703414917,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.6159402274184473,
|
| 1365 |
+
"grad_norm": 0.17195488512516022,
|
| 1366 |
+
"learning_rate": 2.8082259372537635e-05,
|
| 1367 |
+
"loss": 0.08613481521606445,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.6243146696675086,
|
| 1372 |
+
"grad_norm": 0.21867765486240387,
|
| 1373 |
+
"learning_rate": 2.789575792851939e-05,
|
| 1374 |
+
"loss": 0.05220889449119568,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.6326891119165696,
|
| 1379 |
+
"grad_norm": 0.05280565097928047,
|
| 1380 |
+
"learning_rate": 2.770909299132869e-05,
|
| 1381 |
+
"loss": 0.05668648481369019,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.6410635541656307,
|
| 1386 |
+
"grad_norm": 1.2573331594467163,
|
| 1387 |
+
"learning_rate": 2.7522275099982797e-05,
|
| 1388 |
+
"loss": 0.09791213274002075,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.649437996414692,
|
| 1393 |
+
"grad_norm": 0.24661803245544434,
|
| 1394 |
+
"learning_rate": 2.7335314802134676e-05,
|
| 1395 |
+
"loss": 0.05671970248222351,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.6578124386637532,
|
| 1400 |
+
"grad_norm": 0.11609286069869995,
|
| 1401 |
+
"learning_rate": 2.7148222653477505e-05,
|
| 1402 |
+
"loss": 0.052801978588104245,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.6661868809128142,
|
| 1407 |
+
"grad_norm": 0.0745048075914383,
|
| 1408 |
+
"learning_rate": 2.696100921714871e-05,
|
| 1409 |
+
"loss": 0.05944955348968506,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.6745613231618752,
|
| 1414 |
+
"grad_norm": 0.6762334108352661,
|
| 1415 |
+
"learning_rate": 2.677368506313354e-05,
|
| 1416 |
+
"loss": 0.04590807855129242,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.6829357654109365,
|
| 1421 |
+
"grad_norm": 0.0797283723950386,
|
| 1422 |
+
"learning_rate": 2.6586260767668353e-05,
|
| 1423 |
+
"loss": 0.05602290630340576,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.6913102076599977,
|
| 1428 |
+
"grad_norm": 0.0805838331580162,
|
| 1429 |
+
"learning_rate": 2.6398746912643395e-05,
|
| 1430 |
+
"loss": 0.07643119692802429,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.6996846499090588,
|
| 1435 |
+
"grad_norm": 0.5817922353744507,
|
| 1436 |
+
"learning_rate": 2.621115408500545e-05,
|
| 1437 |
+
"loss": 0.06817017793655396,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.7080590921581198,
|
| 1442 |
+
"grad_norm": 0.13826222717761993,
|
| 1443 |
+
"learning_rate": 2.602349287616004e-05,
|
| 1444 |
+
"loss": 0.10364378690719604,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.716433534407181,
|
| 1449 |
+
"grad_norm": 0.7005400061607361,
|
| 1450 |
+
"learning_rate": 2.5835773881373456e-05,
|
| 1451 |
+
"loss": 0.0438168853521347,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.7248079766562423,
|
| 1456 |
+
"grad_norm": 0.213337242603302,
|
| 1457 |
+
"learning_rate": 2.5648007699174566e-05,
|
| 1458 |
+
"loss": 0.07051296830177307,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.7331824189053033,
|
| 1463 |
+
"grad_norm": 0.4417899549007416,
|
| 1464 |
+
"learning_rate": 2.546020493075641e-05,
|
| 1465 |
+
"loss": 0.06489490866661071,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.7415568611543644,
|
| 1470 |
+
"grad_norm": 1.6167367696762085,
|
| 1471 |
+
"learning_rate": 2.527237617937767e-05,
|
| 1472 |
+
"loss": 0.03677745163440704,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.7499313034034256,
|
| 1477 |
+
"grad_norm": 0.834720253944397,
|
| 1478 |
+
"learning_rate": 2.5084532049764027e-05,
|
| 1479 |
+
"loss": 0.05091303586959839,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.7583057456524869,
|
| 1484 |
+
"grad_norm": 0.5853901505470276,
|
| 1485 |
+
"learning_rate": 2.489668314750939e-05,
|
| 1486 |
+
"loss": 0.07355493307113647,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.7666801879015481,
|
| 1491 |
+
"grad_norm": 0.44475457072257996,
|
| 1492 |
+
"learning_rate": 2.4708840078477134e-05,
|
| 1493 |
+
"loss": 0.059920096397399904,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.7750546301506092,
|
| 1498 |
+
"grad_norm": 0.16852276027202606,
|
| 1499 |
+
"learning_rate": 2.4521013448201314e-05,
|
| 1500 |
+
"loss": 0.060352939367294314,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.7834290723996702,
|
| 1505 |
+
"grad_norm": 0.15743663907051086,
|
| 1506 |
+
"learning_rate": 2.433321386128784e-05,
|
| 1507 |
+
"loss": 0.09045843482017517,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.7918035146487314,
|
| 1512 |
+
"grad_norm": 0.4637702405452728,
|
| 1513 |
+
"learning_rate": 2.414545192081575e-05,
|
| 1514 |
+
"loss": 0.08557018041610717,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.8001779568977927,
|
| 1519 |
+
"grad_norm": 0.05946366488933563,
|
| 1520 |
+
"learning_rate": 2.395773822773863e-05,
|
| 1521 |
+
"loss": 0.07715467214584351,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.8085523991468537,
|
| 1526 |
+
"grad_norm": 0.1889827400445938,
|
| 1527 |
+
"learning_rate": 2.3770083380286004e-05,
|
| 1528 |
+
"loss": 0.05025853514671326,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.8169268413959148,
|
| 1533 |
+
"grad_norm": 0.7031142711639404,
|
| 1534 |
+
"learning_rate": 2.3582497973364996e-05,
|
| 1535 |
+
"loss": 0.02963052988052368,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.825301283644976,
|
| 1540 |
+
"grad_norm": 1.4223092794418335,
|
| 1541 |
+
"learning_rate": 2.339499259796216e-05,
|
| 1542 |
+
"loss": 0.06973366737365723,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.8336757258940373,
|
| 1547 |
+
"grad_norm": 0.1039121076464653,
|
| 1548 |
+
"learning_rate": 2.320757784054552e-05,
|
| 1549 |
+
"loss": 0.032325556874275206,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.8420501681430983,
|
| 1554 |
+
"grad_norm": 0.31871646642684937,
|
| 1555 |
+
"learning_rate": 2.302026428246682e-05,
|
| 1556 |
+
"loss": 0.0332093209028244,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.8504246103921593,
|
| 1561 |
+
"grad_norm": 0.22291108965873718,
|
| 1562 |
+
"learning_rate": 2.2833062499364153e-05,
|
| 1563 |
+
"loss": 0.037931859493255615,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.8587990526412206,
|
| 1568 |
+
"grad_norm": 0.5042901635169983,
|
| 1569 |
+
"learning_rate": 2.2645983060564868e-05,
|
| 1570 |
+
"loss": 0.04829805791378021,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.8671734948902818,
|
| 1575 |
+
"grad_norm": 1.4926162958145142,
|
| 1576 |
+
"learning_rate": 2.245903652848877e-05,
|
| 1577 |
+
"loss": 0.04460406899452209,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.8755479371393429,
|
| 1582 |
+
"grad_norm": 0.2670135498046875,
|
| 1583 |
+
"learning_rate": 2.2272233458051837e-05,
|
| 1584 |
+
"loss": 0.045635414123535153,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.883922379388404,
|
| 1589 |
+
"grad_norm": 2.4627184867858887,
|
| 1590 |
+
"learning_rate": 2.2085584396070245e-05,
|
| 1591 |
+
"loss": 0.06845098137855529,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.8922968216374652,
|
| 1596 |
+
"grad_norm": 0.06812036037445068,
|
| 1597 |
+
"learning_rate": 2.1899099880664965e-05,
|
| 1598 |
+
"loss": 0.05972928404808044,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.9006712638865264,
|
| 1603 |
+
"grad_norm": 0.04887200891971588,
|
| 1604 |
+
"learning_rate": 2.17127904406667e-05,
|
| 1605 |
+
"loss": 0.11951442956924438,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.9090457061355874,
|
| 1610 |
+
"grad_norm": 0.2047152817249298,
|
| 1611 |
+
"learning_rate": 2.15266665950215e-05,
|
| 1612 |
+
"loss": 0.023097112774848938,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.9174201483846485,
|
| 1617 |
+
"grad_norm": 0.1777394413948059,
|
| 1618 |
+
"learning_rate": 2.1340738852196834e-05,
|
| 1619 |
+
"loss": 0.04754082262516022,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.9257945906337097,
|
| 1624 |
+
"grad_norm": 0.21407704055309296,
|
| 1625 |
+
"learning_rate": 2.11550177095883e-05,
|
| 1626 |
+
"loss": 0.0496823787689209,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.934169032882771,
|
| 1631 |
+
"grad_norm": 0.06560799479484558,
|
| 1632 |
+
"learning_rate": 2.096951365292693e-05,
|
| 1633 |
+
"loss": 0.04696822464466095,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.942543475131832,
|
| 1638 |
+
"grad_norm": 0.33941805362701416,
|
| 1639 |
+
"learning_rate": 2.0784237155687184e-05,
|
| 1640 |
+
"loss": 0.05176846385002136,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.950917917380893,
|
| 1645 |
+
"grad_norm": 0.32326143980026245,
|
| 1646 |
+
"learning_rate": 2.0599198678495653e-05,
|
| 1647 |
+
"loss": 0.04779183864593506,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.9592923596299543,
|
| 1652 |
+
"grad_norm": 0.10640216618776321,
|
| 1653 |
+
"learning_rate": 2.041440866854038e-05,
|
| 1654 |
+
"loss": 0.03176500201225281,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.9676668018790155,
|
| 1659 |
+
"grad_norm": 0.14410381019115448,
|
| 1660 |
+
"learning_rate": 2.0229877558981075e-05,
|
| 1661 |
+
"loss": 0.10440460443496705,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.9760412441280766,
|
| 1666 |
+
"grad_norm": 0.4526270031929016,
|
| 1667 |
+
"learning_rate": 2.0045615768360078e-05,
|
| 1668 |
+
"loss": 0.03654649257659912,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.9844156863771376,
|
| 1673 |
+
"grad_norm": 0.09464205056428909,
|
| 1674 |
+
"learning_rate": 1.9861633700014065e-05,
|
| 1675 |
+
"loss": 0.041169199347496035,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.9927901286261989,
|
| 1680 |
+
"grad_norm": 0.7187637090682983,
|
| 1681 |
+
"learning_rate": 1.9677941741486735e-05,
|
| 1682 |
+
"loss": 0.07295008301734925,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.000837444224906,
|
| 1687 |
+
"grad_norm": 0.14082716405391693,
|
| 1688 |
+
"learning_rate": 1.949455026394233e-05,
|
| 1689 |
+
"loss": 0.04572053253650665,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.009211886473967,
|
| 1694 |
+
"grad_norm": 2.1813251972198486,
|
| 1695 |
+
"learning_rate": 1.931146962158006e-05,
|
| 1696 |
+
"loss": 0.06506224870681762,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.0175863287230285,
|
| 1701 |
+
"grad_norm": 0.5109630227088928,
|
| 1702 |
+
"learning_rate": 1.912871015104952e-05,
|
| 1703 |
+
"loss": 0.038307076692581175,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.0259607709720897,
|
| 1708 |
+
"grad_norm": 0.5181107521057129,
|
| 1709 |
+
"learning_rate": 1.894628217086709e-05,
|
| 1710 |
+
"loss": 0.04958427548408508,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.0343352132211505,
|
| 1715 |
+
"grad_norm": 0.15236905217170715,
|
| 1716 |
+
"learning_rate": 1.876419598083338e-05,
|
| 1717 |
+
"loss": 0.03732606470584869,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.0427096554702118,
|
| 1722 |
+
"grad_norm": 0.27413368225097656,
|
| 1723 |
+
"learning_rate": 1.8582461861451654e-05,
|
| 1724 |
+
"loss": 0.02974282205104828,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.051084097719273,
|
| 1729 |
+
"grad_norm": 0.30715715885162354,
|
| 1730 |
+
"learning_rate": 1.840109007334743e-05,
|
| 1731 |
+
"loss": 0.03197464346885681,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.0594585399683343,
|
| 1736 |
+
"grad_norm": 0.15932349860668182,
|
| 1737 |
+
"learning_rate": 1.8220090856689192e-05,
|
| 1738 |
+
"loss": 0.020184148848056794,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.067832982217395,
|
| 1743 |
+
"grad_norm": 0.2691262364387512,
|
| 1744 |
+
"learning_rate": 1.8039474430610193e-05,
|
| 1745 |
+
"loss": 0.034737563133239745,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.0762074244664563,
|
| 1750 |
+
"grad_norm": 0.5201924443244934,
|
| 1751 |
+
"learning_rate": 1.7859250992631497e-05,
|
| 1752 |
+
"loss": 0.02623576819896698,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.0845818667155176,
|
| 1757 |
+
"grad_norm": 0.2914315462112427,
|
| 1758 |
+
"learning_rate": 1.7679430718086243e-05,
|
| 1759 |
+
"loss": 0.06491096019744873,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.092956308964579,
|
| 1764 |
+
"grad_norm": 0.1011788621544838,
|
| 1765 |
+
"learning_rate": 1.7500023759545165e-05,
|
| 1766 |
+
"loss": 0.02444133162498474,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.1013307512136397,
|
| 1771 |
+
"grad_norm": 0.4195410907268524,
|
| 1772 |
+
"learning_rate": 1.7321040246243335e-05,
|
| 1773 |
+
"loss": 0.03377276659011841,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.109705193462701,
|
| 1778 |
+
"grad_norm": 0.2820679843425751,
|
| 1779 |
+
"learning_rate": 1.7142490283508324e-05,
|
| 1780 |
+
"loss": 0.038240480422973636,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.118079635711762,
|
| 1785 |
+
"grad_norm": 0.1285504251718521,
|
| 1786 |
+
"learning_rate": 1.696438395218963e-05,
|
| 1787 |
+
"loss": 0.04785881340503693,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.1264540779608234,
|
| 1792 |
+
"grad_norm": 0.07012035697698593,
|
| 1793 |
+
"learning_rate": 1.6786731308089513e-05,
|
| 1794 |
+
"loss": 0.035026317834854125,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.1348285202098847,
|
| 1799 |
+
"grad_norm": 0.1849839687347412,
|
| 1800 |
+
"learning_rate": 1.6609542381395266e-05,
|
| 1801 |
+
"loss": 0.019530372321605684,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.1432029624589455,
|
| 1806 |
+
"grad_norm": 0.19696128368377686,
|
| 1807 |
+
"learning_rate": 1.64328271761129e-05,
|
| 1808 |
+
"loss": 0.031582942605018614,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.1515774047080067,
|
| 1813 |
+
"grad_norm": 0.575410783290863,
|
| 1814 |
+
"learning_rate": 1.625659566950236e-05,
|
| 1815 |
+
"loss": 0.06284295320510865,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.159951846957068,
|
| 1820 |
+
"grad_norm": 0.26818394660949707,
|
| 1821 |
+
"learning_rate": 1.6080857811514148e-05,
|
| 1822 |
+
"loss": 0.033028793334960935,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.1683262892061292,
|
| 1827 |
+
"grad_norm": 0.4050503671169281,
|
| 1828 |
+
"learning_rate": 1.5905623524227616e-05,
|
| 1829 |
+
"loss": 0.027020111680030823,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.17670073145519,
|
| 1834 |
+
"grad_norm": 0.9115064144134521,
|
| 1835 |
+
"learning_rate": 1.573090270129072e-05,
|
| 1836 |
+
"loss": 0.037383803725242616,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.1850751737042513,
|
| 1841 |
+
"grad_norm": 0.055088263005018234,
|
| 1842 |
+
"learning_rate": 1.555670520736149e-05,
|
| 1843 |
+
"loss": 0.04163878262042999,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.1934496159533126,
|
| 1848 |
+
"grad_norm": 0.08490361273288727,
|
| 1849 |
+
"learning_rate": 1.5383040877550997e-05,
|
| 1850 |
+
"loss": 0.027175077795982362,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.201824058202374,
|
| 1855 |
+
"grad_norm": 0.05155947431921959,
|
| 1856 |
+
"learning_rate": 1.5209919516868129e-05,
|
| 1857 |
+
"loss": 0.017401106655597687,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.2101985004514346,
|
| 1862 |
+
"grad_norm": 0.3288068473339081,
|
| 1863 |
+
"learning_rate": 1.5037350899665992e-05,
|
| 1864 |
+
"loss": 0.0302372545003891,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.218572942700496,
|
| 1869 |
+
"grad_norm": 0.16062849760055542,
|
| 1870 |
+
"learning_rate": 1.4865344769090034e-05,
|
| 1871 |
+
"loss": 0.039090073108673094,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.226947384949557,
|
| 1876 |
+
"grad_norm": 0.39235061407089233,
|
| 1877 |
+
"learning_rate": 1.4693910836527956e-05,
|
| 1878 |
+
"loss": 0.04479800760746002,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.2353218271986184,
|
| 1883 |
+
"grad_norm": 0.08488383889198303,
|
| 1884 |
+
"learning_rate": 1.4523058781061454e-05,
|
| 1885 |
+
"loss": 0.03554574549198151,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.243696269447679,
|
| 1890 |
+
"grad_norm": 0.1608286052942276,
|
| 1891 |
+
"learning_rate": 1.435279824891968e-05,
|
| 1892 |
+
"loss": 0.04044436812400818,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.2520707116967404,
|
| 1897 |
+
"grad_norm": 0.10187438875436783,
|
| 1898 |
+
"learning_rate": 1.4183138852934641e-05,
|
| 1899 |
+
"loss": 0.041433611512184144,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.2604451539458017,
|
| 1904 |
+
"grad_norm": 1.007249116897583,
|
| 1905 |
+
"learning_rate": 1.4014090171998511e-05,
|
| 1906 |
+
"loss": 0.03556722104549408,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.268819596194863,
|
| 1911 |
+
"grad_norm": 0.1751636415719986,
|
| 1912 |
+
"learning_rate": 1.3845661750522733e-05,
|
| 1913 |
+
"loss": 0.02698090374469757,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.2771940384439238,
|
| 1918 |
+
"grad_norm": 0.5745561122894287,
|
| 1919 |
+
"learning_rate": 1.3677863097899196e-05,
|
| 1920 |
+
"loss": 0.027302804589271545,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.285568480692985,
|
| 1925 |
+
"grad_norm": 0.18385080993175507,
|
| 1926 |
+
"learning_rate": 1.3510703687963317e-05,
|
| 1927 |
+
"loss": 0.036330342292785645,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.2939429229420463,
|
| 1932 |
+
"grad_norm": 0.8466984629631042,
|
| 1933 |
+
"learning_rate": 1.334419295845919e-05,
|
| 1934 |
+
"loss": 0.04368994832038879,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.3023173651911075,
|
| 1939 |
+
"grad_norm": 0.5879298448562622,
|
| 1940 |
+
"learning_rate": 1.3178340310506681e-05,
|
| 1941 |
+
"loss": 0.0322884202003479,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.3106918074401683,
|
| 1946 |
+
"grad_norm": 0.1511489599943161,
|
| 1947 |
+
"learning_rate": 1.3013155108070685e-05,
|
| 1948 |
+
"loss": 0.05695425868034363,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.3190662496892296,
|
| 1953 |
+
"grad_norm": 0.059570491313934326,
|
| 1954 |
+
"learning_rate": 1.2848646677432416e-05,
|
| 1955 |
+
"loss": 0.02256210893392563,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.327440691938291,
|
| 1960 |
+
"grad_norm": 0.0460665300488472,
|
| 1961 |
+
"learning_rate": 1.2684824306662871e-05,
|
| 1962 |
+
"loss": 0.016327956318855287,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.335815134187352,
|
| 1967 |
+
"grad_norm": 0.05761329084634781,
|
| 1968 |
+
"learning_rate": 1.2521697245098413e-05,
|
| 1969 |
+
"loss": 0.030600738525390626,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.3441895764364133,
|
| 1974 |
+
"grad_norm": 0.07144748419523239,
|
| 1975 |
+
"learning_rate": 1.2359274702818564e-05,
|
| 1976 |
+
"loss": 0.0311413437128067,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.352564018685474,
|
| 1981 |
+
"grad_norm": 0.05444992706179619,
|
| 1982 |
+
"learning_rate": 1.2197565850126023e-05,
|
| 1983 |
+
"loss": 0.026527315378189087,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.3609384609345354,
|
| 1988 |
+
"grad_norm": 0.39611950516700745,
|
| 1989 |
+
"learning_rate": 1.2036579817028884e-05,
|
| 1990 |
+
"loss": 0.035798245668411256,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.3693129031835967,
|
| 1995 |
+
"grad_norm": 0.03898654505610466,
|
| 1996 |
+
"learning_rate": 1.1876325692725171e-05,
|
| 1997 |
+
"loss": 0.03876761496067047,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.3776873454326575,
|
| 2002 |
+
"grad_norm": 0.06425919383764267,
|
| 2003 |
+
"learning_rate": 1.1716812525089702e-05,
|
| 2004 |
+
"loss": 0.025335514545440675,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.3860617876817187,
|
| 2009 |
+
"grad_norm": 0.07012301683425903,
|
| 2010 |
+
"learning_rate": 1.1558049320163186e-05,
|
| 2011 |
+
"loss": 0.0288514643907547,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.39443622993078,
|
| 2016 |
+
"grad_norm": 0.12998658418655396,
|
| 2017 |
+
"learning_rate": 1.1400045041643795e-05,
|
| 2018 |
+
"loss": 0.04393172860145569,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.4028106721798412,
|
| 2023 |
+
"grad_norm": 0.4555356204509735,
|
| 2024 |
+
"learning_rate": 1.1242808610381051e-05,
|
| 2025 |
+
"loss": 0.034080561995506284,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.4111851144289025,
|
| 2030 |
+
"grad_norm": 0.6439165472984314,
|
| 2031 |
+
"learning_rate": 1.108634890387219e-05,
|
| 2032 |
+
"loss": 0.019711557030677795,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.4195595566779633,
|
| 2037 |
+
"grad_norm": 0.04977191612124443,
|
| 2038 |
+
"learning_rate": 1.0930674755760908e-05,
|
| 2039 |
+
"loss": 0.044436204433441165,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.4279339989270246,
|
| 2044 |
+
"grad_norm": 0.04331052675843239,
|
| 2045 |
+
"learning_rate": 1.0775794955338625e-05,
|
| 2046 |
+
"loss": 0.038460010290145875,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.436308441176086,
|
| 2051 |
+
"grad_norm": 0.5490692853927612,
|
| 2052 |
+
"learning_rate": 1.0621718247048278e-05,
|
| 2053 |
+
"loss": 0.030916082859039306,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.444682883425147,
|
| 2058 |
+
"grad_norm": 0.17242658138275146,
|
| 2059 |
+
"learning_rate": 1.0468453329990563e-05,
|
| 2060 |
+
"loss": 0.025243097543716432,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.453057325674208,
|
| 2065 |
+
"grad_norm": 0.3855954706668854,
|
| 2066 |
+
"learning_rate": 1.0316008857432827e-05,
|
| 2067 |
+
"loss": 0.030227524042129517,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.461431767923269,
|
| 2072 |
+
"grad_norm": 0.564544677734375,
|
| 2073 |
+
"learning_rate": 1.0164393436320477e-05,
|
| 2074 |
+
"loss": 0.03143407106399536,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.4698062101723304,
|
| 2079 |
+
"grad_norm": 0.31516560912132263,
|
| 2080 |
+
"learning_rate": 1.0013615626791087e-05,
|
| 2081 |
+
"loss": 0.035627859830856326,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.4781806524213916,
|
| 2086 |
+
"grad_norm": 0.06647393107414246,
|
| 2087 |
+
"learning_rate": 9.863683941691033e-06,
|
| 2088 |
+
"loss": 0.03705155551433563,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.4865550946704524,
|
| 2093 |
+
"grad_norm": 0.2980024814605713,
|
| 2094 |
+
"learning_rate": 9.714606846094892e-06,
|
| 2095 |
+
"loss": 0.05262235403060913,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.4949295369195137,
|
| 2100 |
+
"grad_norm": 0.05878954753279686,
|
| 2101 |
+
"learning_rate": 9.566392756827533e-06,
|
| 2102 |
+
"loss": 0.019335083663463593,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.503303979168575,
|
| 2107 |
+
"grad_norm": 0.5564544796943665,
|
| 2108 |
+
"learning_rate": 9.419050041988859e-06,
|
| 2109 |
+
"loss": 0.028229591250419617,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.511678421417636,
|
| 2114 |
+
"grad_norm": 0.08901913464069366,
|
| 2115 |
+
"learning_rate": 9.272587020481371e-06,
|
| 2116 |
+
"loss": 0.03342890441417694,
|
| 2117 |
+
"step": 6000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.520052863666697,
|
| 2121 |
+
"grad_norm": 0.8243499994277954,
|
| 2122 |
+
"learning_rate": 9.127011961540491e-06,
|
| 2123 |
+
"loss": 0.021094033122062684,
|
| 2124 |
+
"step": 6020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.5284273059157583,
|
| 2128 |
+
"grad_norm": 0.17781858146190643,
|
| 2129 |
+
"learning_rate": 8.982333084267683e-06,
|
| 2130 |
+
"loss": 0.029926052689552306,
|
| 2131 |
+
"step": 6040
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.5368017481648195,
|
| 2135 |
+
"grad_norm": 0.04101826995611191,
|
| 2136 |
+
"learning_rate": 8.838558557166391e-06,
|
| 2137 |
+
"loss": 0.04929870367050171,
|
| 2138 |
+
"step": 6060
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.5451761904138808,
|
| 2142 |
+
"grad_norm": 0.310251384973526,
|
| 2143 |
+
"learning_rate": 8.695696497680864e-06,
|
| 2144 |
+
"loss": 0.027747470140457153,
|
| 2145 |
+
"step": 6080
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.553550632662942,
|
| 2149 |
+
"grad_norm": 0.724092960357666,
|
| 2150 |
+
"learning_rate": 8.553754971737864e-06,
|
| 2151 |
+
"loss": 0.03592235445976257,
|
| 2152 |
+
"step": 6100
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.561925074912003,
|
| 2156 |
+
"grad_norm": 0.21852202713489532,
|
| 2157 |
+
"learning_rate": 8.412741993291226e-06,
|
| 2158 |
+
"loss": 0.03389221727848053,
|
| 2159 |
+
"step": 6120
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.570299517161064,
|
| 2163 |
+
"grad_norm": 0.3016412854194641,
|
| 2164 |
+
"learning_rate": 8.272665523869416e-06,
|
| 2165 |
+
"loss": 0.04466528296470642,
|
| 2166 |
+
"step": 6140
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.5786739594101253,
|
| 2170 |
+
"grad_norm": 0.3840382695198059,
|
| 2171 |
+
"learning_rate": 8.13353347212605e-06,
|
| 2172 |
+
"loss": 0.036823111772537234,
|
| 2173 |
+
"step": 6160
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.587048401659186,
|
| 2177 |
+
"grad_norm": 0.052707478404045105,
|
| 2178 |
+
"learning_rate": 7.995353693393337e-06,
|
| 2179 |
+
"loss": 0.030457329750061036,
|
| 2180 |
+
"step": 6180
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.5954228439082474,
|
| 2184 |
+
"grad_norm": 0.861706554889679,
|
| 2185 |
+
"learning_rate": 7.85813398923858e-06,
|
| 2186 |
+
"loss": 0.0386508047580719,
|
| 2187 |
+
"step": 6200
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.6037972861573087,
|
| 2191 |
+
"grad_norm": 0.036389291286468506,
|
| 2192 |
+
"learning_rate": 7.721882107023711e-06,
|
| 2193 |
+
"loss": 0.022716432809829712,
|
| 2194 |
+
"step": 6220
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.61217172840637,
|
| 2198 |
+
"grad_norm": 0.5068594813346863,
|
| 2199 |
+
"learning_rate": 7.5866057394679e-06,
|
| 2200 |
+
"loss": 0.037295091152191165,
|
| 2201 |
+
"step": 6240
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.620546170655431,
|
| 2205 |
+
"grad_norm": 0.03255036100745201,
|
| 2206 |
+
"learning_rate": 7.452312524213173e-06,
|
| 2207 |
+
"loss": 0.023309530317783357,
|
| 2208 |
+
"step": 6260
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.628920612904492,
|
| 2212 |
+
"grad_norm": 0.37486711144447327,
|
| 2213 |
+
"learning_rate": 7.319010043393235e-06,
|
| 2214 |
+
"loss": 0.02240334004163742,
|
| 2215 |
+
"step": 6280
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.6372950551535532,
|
| 2219 |
+
"grad_norm": 0.17838098108768463,
|
| 2220 |
+
"learning_rate": 7.186705823205394e-06,
|
| 2221 |
+
"loss": 0.021342363953590394,
|
| 2222 |
+
"step": 6300
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.6456694974026145,
|
| 2226 |
+
"grad_norm": 0.0714629665017128,
|
| 2227 |
+
"learning_rate": 7.055407333485594e-06,
|
| 2228 |
+
"loss": 0.027359399199485778,
|
| 2229 |
+
"step": 6320
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.6540439396516753,
|
| 2233 |
+
"grad_norm": 0.49168136715888977,
|
| 2234 |
+
"learning_rate": 6.925121987286709e-06,
|
| 2235 |
+
"loss": 0.12207809686660767,
|
| 2236 |
+
"step": 6340
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.6624183819007365,
|
| 2240 |
+
"grad_norm": 0.13302330672740936,
|
| 2241 |
+
"learning_rate": 6.7958571404599766e-06,
|
| 2242 |
+
"loss": 0.02340977340936661,
|
| 2243 |
+
"step": 6360
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.670792824149798,
|
| 2247 |
+
"grad_norm": 0.22196047008037567,
|
| 2248 |
+
"learning_rate": 6.667620091239732e-06,
|
| 2249 |
+
"loss": 0.02150045782327652,
|
| 2250 |
+
"step": 6380
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.679167266398859,
|
| 2254 |
+
"grad_norm": 0.05339694768190384,
|
| 2255 |
+
"learning_rate": 6.540418079831301e-06,
|
| 2256 |
+
"loss": 0.04855234920978546,
|
| 2257 |
+
"step": 6400
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.6875417086479203,
|
| 2261 |
+
"grad_norm": 0.04468238353729248,
|
| 2262 |
+
"learning_rate": 6.414258288002259e-06,
|
| 2263 |
+
"loss": 0.01810908317565918,
|
| 2264 |
+
"step": 6420
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.695916150896981,
|
| 2268 |
+
"grad_norm": 0.17243818938732147,
|
| 2269 |
+
"learning_rate": 6.2891478386769315e-06,
|
| 2270 |
+
"loss": 0.02477678209543228,
|
| 2271 |
+
"step": 6440
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.7042905931460424,
|
| 2275 |
+
"grad_norm": 0.11534826457500458,
|
| 2276 |
+
"learning_rate": 6.165093795534254e-06,
|
| 2277 |
+
"loss": 0.026881667971611022,
|
| 2278 |
+
"step": 6460
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.7126650353951036,
|
| 2282 |
+
"grad_norm": 0.36479777097702026,
|
| 2283 |
+
"learning_rate": 6.042103162608937e-06,
|
| 2284 |
+
"loss": 0.021916355192661285,
|
| 2285 |
+
"step": 6480
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.721039477644165,
|
| 2289 |
+
"grad_norm": 0.3529409170150757,
|
| 2290 |
+
"learning_rate": 5.920182883896053e-06,
|
| 2291 |
+
"loss": 0.022671876847743987,
|
| 2292 |
+
"step": 6500
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.7294139198932257,
|
| 2296 |
+
"grad_norm": 0.24025386571884155,
|
| 2297 |
+
"learning_rate": 5.7993398429589506e-06,
|
| 2298 |
+
"loss": 0.05065078735351562,
|
| 2299 |
+
"step": 6520
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.737788362142287,
|
| 2303 |
+
"grad_norm": 0.13825002312660217,
|
| 2304 |
+
"learning_rate": 5.679580862540629e-06,
|
| 2305 |
+
"loss": 0.06285998821258545,
|
| 2306 |
+
"step": 6540
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.746162804391348,
|
| 2310 |
+
"grad_norm": 0.09770816564559937,
|
| 2311 |
+
"learning_rate": 5.560912704178514e-06,
|
| 2312 |
+
"loss": 0.1244073748588562,
|
| 2313 |
+
"step": 6560
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.7545372466404094,
|
| 2317 |
+
"grad_norm": 0.3843393623828888,
|
| 2318 |
+
"learning_rate": 5.443342067822743e-06,
|
| 2319 |
+
"loss": 0.018991626799106598,
|
| 2320 |
+
"step": 6580
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.7629116888894707,
|
| 2324 |
+
"grad_norm": 0.34221217036247253,
|
| 2325 |
+
"learning_rate": 5.326875591457833e-06,
|
| 2326 |
+
"loss": 0.025901442766189574,
|
| 2327 |
+
"step": 6600
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.7712861311385315,
|
| 2331 |
+
"grad_norm": 0.07205979526042938,
|
| 2332 |
+
"learning_rate": 5.2115198507279365e-06,
|
| 2333 |
+
"loss": 0.02461903393268585,
|
| 2334 |
+
"step": 6620
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.7796605733875928,
|
| 2338 |
+
"grad_norm": 0.16574513912200928,
|
| 2339 |
+
"learning_rate": 5.097281358565595e-06,
|
| 2340 |
+
"loss": 0.030345502495765685,
|
| 2341 |
+
"step": 6640
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.788035015636654,
|
| 2345 |
+
"grad_norm": 0.6843248605728149,
|
| 2346 |
+
"learning_rate": 4.984166564823992e-06,
|
| 2347 |
+
"loss": 0.03047039806842804,
|
| 2348 |
+
"step": 6660
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.796409457885715,
|
| 2352 |
+
"grad_norm": 0.08679454028606415,
|
| 2353 |
+
"learning_rate": 4.872181855912811e-06,
|
| 2354 |
+
"loss": 0.03248770534992218,
|
| 2355 |
+
"step": 6680
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.804783900134776,
|
| 2359 |
+
"grad_norm": 0.038589492440223694,
|
| 2360 |
+
"learning_rate": 4.761333554437653e-06,
|
| 2361 |
+
"loss": 0.026063334941864014,
|
| 2362 |
+
"step": 6700
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.8131583423838373,
|
| 2366 |
+
"grad_norm": 0.6347565650939941,
|
| 2367 |
+
"learning_rate": 4.651627918843099e-06,
|
| 2368 |
+
"loss": 0.02364843189716339,
|
| 2369 |
+
"step": 6720
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.8215327846328986,
|
| 2373 |
+
"grad_norm": 0.03580786660313606,
|
| 2374 |
+
"learning_rate": 4.543071143059311e-06,
|
| 2375 |
+
"loss": 0.02118242233991623,
|
| 2376 |
+
"step": 6740
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.82990722688196,
|
| 2380 |
+
"grad_norm": 0.6818614602088928,
|
| 2381 |
+
"learning_rate": 4.435669356152352e-06,
|
| 2382 |
+
"loss": 0.026198378205299376,
|
| 2383 |
+
"step": 6760
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.8382816691310206,
|
| 2387 |
+
"grad_norm": 0.18815983831882477,
|
| 2388 |
+
"learning_rate": 4.3294286219781595e-06,
|
| 2389 |
+
"loss": 0.029299169778823853,
|
| 2390 |
+
"step": 6780
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.846656111380082,
|
| 2394 |
+
"grad_norm": 0.49268460273742676,
|
| 2395 |
+
"learning_rate": 4.224354938840144e-06,
|
| 2396 |
+
"loss": 0.03095126748085022,
|
| 2397 |
+
"step": 6800
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.855030553629143,
|
| 2401 |
+
"grad_norm": 0.15540994703769684,
|
| 2402 |
+
"learning_rate": 4.120454239150548e-06,
|
| 2403 |
+
"loss": 0.024085646867752074,
|
| 2404 |
+
"step": 6820
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.863404995878204,
|
| 2408 |
+
"grad_norm": 0.03290439397096634,
|
| 2409 |
+
"learning_rate": 4.017732389095511e-06,
|
| 2410 |
+
"loss": 0.02222718745470047,
|
| 2411 |
+
"step": 6840
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.871779438127265,
|
| 2415 |
+
"grad_norm": 0.25523442029953003,
|
| 2416 |
+
"learning_rate": 3.916195188303853e-06,
|
| 2417 |
+
"loss": 0.03877650797367096,
|
| 2418 |
+
"step": 6860
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 2.8801538803763265,
|
| 2422 |
+
"grad_norm": 0.06997464597225189,
|
| 2423 |
+
"learning_rate": 3.81584836951964e-06,
|
| 2424 |
+
"loss": 0.04235384464263916,
|
| 2425 |
+
"step": 6880
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 2.8885283226253877,
|
| 2429 |
+
"grad_norm": 0.13868676126003265,
|
| 2430 |
+
"learning_rate": 3.716697598278507e-06,
|
| 2431 |
+
"loss": 0.031541892886161806,
|
| 2432 |
+
"step": 6900
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 2.896902764874449,
|
| 2436 |
+
"grad_norm": 0.023875195533037186,
|
| 2437 |
+
"learning_rate": 3.6187484725878047e-06,
|
| 2438 |
+
"loss": 0.01966340243816376,
|
| 2439 |
+
"step": 6920
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.90527720712351,
|
| 2443 |
+
"grad_norm": 0.06577973067760468,
|
| 2444 |
+
"learning_rate": 3.5220065226105105e-06,
|
| 2445 |
+
"loss": 0.02548326551914215,
|
| 2446 |
+
"step": 6940
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.913651649372571,
|
| 2450 |
+
"grad_norm": 0.11466669291257858,
|
| 2451 |
+
"learning_rate": 3.426477210353013e-06,
|
| 2452 |
+
"loss": 0.026891687512397768,
|
| 2453 |
+
"step": 6960
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.9220260916216323,
|
| 2457 |
+
"grad_norm": 0.11031544208526611,
|
| 2458 |
+
"learning_rate": 3.332165929356726e-06,
|
| 2459 |
+
"loss": 0.020203272998332977,
|
| 2460 |
+
"step": 6980
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 2.930400533870693,
|
| 2464 |
+
"grad_norm": 0.6550182700157166,
|
| 2465 |
+
"learning_rate": 3.2390780043935864e-06,
|
| 2466 |
+
"loss": 0.04080580472946167,
|
| 2467 |
+
"step": 7000
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 2.9387749761197544,
|
| 2471 |
+
"grad_norm": 0.06129005551338196,
|
| 2472 |
+
"learning_rate": 3.147218691165396e-06,
|
| 2473 |
+
"loss": 0.01906572878360748,
|
| 2474 |
+
"step": 7020
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 2.9471494183688156,
|
| 2478 |
+
"grad_norm": 0.37061557173728943,
|
| 2479 |
+
"learning_rate": 3.0565931760070917e-06,
|
| 2480 |
+
"loss": 0.02230120003223419,
|
| 2481 |
+
"step": 7040
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.955523860617877,
|
| 2485 |
+
"grad_norm": 0.11101196706295013,
|
| 2486 |
+
"learning_rate": 2.967206575593956e-06,
|
| 2487 |
+
"loss": 0.0184899240732193,
|
| 2488 |
+
"step": 7060
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 2.963898302866938,
|
| 2492 |
+
"grad_norm": 0.06029270961880684,
|
| 2493 |
+
"learning_rate": 2.87906393665269e-06,
|
| 2494 |
+
"loss": 0.030630800127983093,
|
| 2495 |
+
"step": 7080
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 2.9722727451159994,
|
| 2499 |
+
"grad_norm": 0.1466781049966812,
|
| 2500 |
+
"learning_rate": 2.792170235676497e-06,
|
| 2501 |
+
"loss": 0.030768883228302003,
|
| 2502 |
+
"step": 7100
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.98064718736506,
|
| 2506 |
+
"grad_norm": 0.6590607762336731,
|
| 2507 |
+
"learning_rate": 2.7065303786441314e-06,
|
| 2508 |
+
"loss": 0.025333571434020995,
|
| 2509 |
+
"step": 7120
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 2.9890216296141214,
|
| 2513 |
+
"grad_norm": 0.04609564691781998,
|
| 2514 |
+
"learning_rate": 2.622149200742871e-06,
|
| 2515 |
+
"loss": 0.028621068596839903,
|
| 2516 |
+
"step": 7140
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.9973960718631827,
|
| 2520 |
+
"grad_norm": 0.05696053057909012,
|
| 2521 |
+
"learning_rate": 2.539031466095551e-06,
|
| 2522 |
+
"loss": 0.030104705691337587,
|
| 2523 |
+
"step": 7160
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 3.0054433874618898,
|
| 2527 |
+
"grad_norm": 0.08858000487089157,
|
| 2528 |
+
"learning_rate": 2.457181867491573e-06,
|
| 2529 |
+
"loss": 0.014444483816623688,
|
| 2530 |
+
"step": 7180
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 3.013817829710951,
|
| 2534 |
+
"grad_norm": 0.8151156902313232,
|
| 2535 |
+
"learning_rate": 2.376605026121964e-06,
|
| 2536 |
+
"loss": 0.030318111181259155,
|
| 2537 |
+
"step": 7200
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 3.022192271960012,
|
| 2541 |
+
"grad_norm": 0.08565035462379456,
|
| 2542 |
+
"learning_rate": 2.297305491318447e-06,
|
| 2543 |
+
"loss": 0.01848241835832596,
|
| 2544 |
+
"step": 7220
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 3.030566714209073,
|
| 2548 |
+
"grad_norm": 0.02643604762852192,
|
| 2549 |
+
"learning_rate": 2.219287740296605e-06,
|
| 2550 |
+
"loss": 0.023431843519210814,
|
| 2551 |
+
"step": 7240
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 3.0389411564581343,
|
| 2555 |
+
"grad_norm": 0.038074638694524765,
|
| 2556 |
+
"learning_rate": 2.142556177903096e-06,
|
| 2557 |
+
"loss": 0.02755042314529419,
|
| 2558 |
+
"step": 7260
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 3.0473155987071956,
|
| 2562 |
+
"grad_norm": 0.03524810075759888,
|
| 2563 |
+
"learning_rate": 2.067115136366948e-06,
|
| 2564 |
+
"loss": 0.03158809542655945,
|
| 2565 |
+
"step": 7280
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 3.0556900409562564,
|
| 2569 |
+
"grad_norm": 0.0943976640701294,
|
| 2570 |
+
"learning_rate": 1.9929688750549734e-06,
|
| 2571 |
+
"loss": 0.016621415317058564,
|
| 2572 |
+
"step": 7300
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 3.0640644832053177,
|
| 2576 |
+
"grad_norm": 0.026402320712804794,
|
| 2577 |
+
"learning_rate": 1.9201215802312794e-06,
|
| 2578 |
+
"loss": 0.029634922742843628,
|
| 2579 |
+
"step": 7320
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 2.3046213333071632,
|
| 2583 |
+
"grad_norm": 20.98118782043457,
|
| 2584 |
+
"learning_rate": 1.8485773648209192e-06,
|
| 2585 |
+
"loss": 17.70731715152138,
|
| 2586 |
+
"step": 7340
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 2.3109021855403005,
|
| 2590 |
+
"grad_norm": 23.614276885986328,
|
| 2591 |
+
"learning_rate": 1.7783402681776833e-06,
|
| 2592 |
+
"loss": 21.242208862304686,
|
| 2593 |
+
"step": 7360
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 2.3171830377734377,
|
| 2597 |
+
"grad_norm": 16.01496124267578,
|
| 2598 |
+
"learning_rate": 1.7094142558560177e-06,
|
| 2599 |
+
"loss": 16.543919372558594,
|
| 2600 |
+
"step": 7380
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 2.3234638900065754,
|
| 2604 |
+
"grad_norm": 27.359785079956055,
|
| 2605 |
+
"learning_rate": 1.6418032193871612e-06,
|
| 2606 |
+
"loss": 17.941021728515626,
|
| 2607 |
+
"step": 7400
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.3297447422397126,
|
| 2611 |
+
"grad_norm": 21.860763549804688,
|
| 2612 |
+
"learning_rate": 1.5755109760594105e-06,
|
| 2613 |
+
"loss": 17.183251953125,
|
| 2614 |
+
"step": 7420
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 2.33602559447285,
|
| 2618 |
+
"grad_norm": 27.401737213134766,
|
| 2619 |
+
"learning_rate": 1.510541268702595e-06,
|
| 2620 |
+
"loss": 17.2944091796875,
|
| 2621 |
+
"step": 7440
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.3423064467059875,
|
| 2625 |
+
"grad_norm": 18.479413986206055,
|
| 2626 |
+
"learning_rate": 1.4468977654767736e-06,
|
| 2627 |
+
"loss": 16.43244323730469,
|
| 2628 |
+
"step": 7460
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 2.3485872989391248,
|
| 2632 |
+
"grad_norm": 16.50922203063965,
|
| 2633 |
+
"learning_rate": 1.3845840596651256e-06,
|
| 2634 |
+
"loss": 14.042901611328125,
|
| 2635 |
+
"step": 7480
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 2.354868151172262,
|
| 2639 |
+
"grad_norm": 18.05698013305664,
|
| 2640 |
+
"learning_rate": 1.3236036694710712e-06,
|
| 2641 |
+
"loss": 13.512533569335938,
|
| 2642 |
+
"step": 7500
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 2.3611490034053997,
|
| 2646 |
+
"grad_norm": 23.549781799316406,
|
| 2647 |
+
"learning_rate": 1.2639600378196336e-06,
|
| 2648 |
+
"loss": 15.484471130371094,
|
| 2649 |
+
"step": 7520
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 2.367429855638537,
|
| 2653 |
+
"grad_norm": 34.259979248046875,
|
| 2654 |
+
"learning_rate": 1.2056565321630742e-06,
|
| 2655 |
+
"loss": 14.65059814453125,
|
| 2656 |
+
"step": 7540
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 2.373710707871674,
|
| 2660 |
+
"grad_norm": 21.186641693115234,
|
| 2661 |
+
"learning_rate": 1.1486964442907338e-06,
|
| 2662 |
+
"loss": 14.287913513183593,
|
| 2663 |
+
"step": 7560
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 2.379991560104812,
|
| 2667 |
+
"grad_norm": 23.37986946105957,
|
| 2668 |
+
"learning_rate": 1.093082990143207e-06,
|
| 2669 |
+
"loss": 14.952766418457031,
|
| 2670 |
+
"step": 7580
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 2.386272412337949,
|
| 2674 |
+
"grad_norm": 32.49317932128906,
|
| 2675 |
+
"learning_rate": 1.0388193096307619e-06,
|
| 2676 |
+
"loss": 14.402635192871093,
|
| 2677 |
+
"step": 7600
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 2.3925532645710863,
|
| 2681 |
+
"grad_norm": 27.62471580505371,
|
| 2682 |
+
"learning_rate": 9.85908466456062e-07,
|
| 2683 |
+
"loss": 13.837322998046876,
|
| 2684 |
+
"step": 7620
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 2.398834116804224,
|
| 2688 |
+
"grad_norm": 36.663368225097656,
|
| 2689 |
+
"learning_rate": 9.343534479411886e-07,
|
| 2690 |
+
"loss": 13.763095092773437,
|
| 2691 |
+
"step": 7640
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 2.405114969037361,
|
| 2695 |
+
"grad_norm": 20.94727325439453,
|
| 2696 |
+
"learning_rate": 8.841571648589814e-07,
|
| 2697 |
+
"loss": 12.785398864746094,
|
| 2698 |
+
"step": 7660
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 2.4113958212704985,
|
| 2702 |
+
"grad_norm": 13.804335594177246,
|
| 2703 |
+
"learning_rate": 8.353224512686997e-07,
|
| 2704 |
+
"loss": 14.701901245117188,
|
| 2705 |
+
"step": 7680
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 2.417676673503636,
|
| 2709 |
+
"grad_norm": 41.34111022949219,
|
| 2710 |
+
"learning_rate": 7.878520643560123e-07,
|
| 2711 |
+
"loss": 14.57025146484375,
|
| 2712 |
+
"step": 7700
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 2.4239575257367734,
|
| 2716 |
+
"grad_norm": 32.85154724121094,
|
| 2717 |
+
"learning_rate": 7.417486842773186e-07,
|
| 2718 |
+
"loss": 13.9228271484375,
|
| 2719 |
+
"step": 7720
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 2.4302383779699106,
|
| 2723 |
+
"grad_norm": 20.45840835571289,
|
| 2724 |
+
"learning_rate": 6.970149140084398e-07,
|
| 2725 |
+
"loss": 13.153306579589843,
|
| 2726 |
+
"step": 7740
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 2.4365192302030483,
|
| 2730 |
+
"grad_norm": 22.2620792388916,
|
| 2731 |
+
"learning_rate": 6.536532791976496e-07,
|
| 2732 |
+
"loss": 13.955767822265624,
|
| 2733 |
+
"step": 7760
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 2.4428000824361855,
|
| 2737 |
+
"grad_norm": 19.04249382019043,
|
| 2738 |
+
"learning_rate": 6.116662280230833e-07,
|
| 2739 |
+
"loss": 13.429383850097656,
|
| 2740 |
+
"step": 7780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 2.4490809346693228,
|
| 2744 |
+
"grad_norm": 21.958354949951172,
|
| 2745 |
+
"learning_rate": 5.710561310545037e-07,
|
| 2746 |
+
"loss": 14.128433227539062,
|
| 2747 |
+
"step": 7800
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 2.4553617869024604,
|
| 2751 |
+
"grad_norm": 12.351365089416504,
|
| 2752 |
+
"learning_rate": 5.318252811194773e-07,
|
| 2753 |
+
"loss": 13.092503356933594,
|
| 2754 |
+
"step": 7820
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 2.4616426391355977,
|
| 2758 |
+
"grad_norm": 16.454404830932617,
|
| 2759 |
+
"learning_rate": 4.939758931739036e-07,
|
| 2760 |
+
"loss": 14.190911865234375,
|
| 2761 |
+
"step": 7840
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 2.467923491368735,
|
| 2765 |
+
"grad_norm": 18.192588806152344,
|
| 2766 |
+
"learning_rate": 4.5751010417696417e-07,
|
| 2767 |
+
"loss": 12.131760406494141,
|
| 2768 |
+
"step": 7860
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 2.4742043436018726,
|
| 2772 |
+
"grad_norm": 26.641132354736328,
|
| 2773 |
+
"learning_rate": 4.2242997297049224e-07,
|
| 2774 |
+
"loss": 13.599835205078126,
|
| 2775 |
+
"step": 7880
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 2.48048519583501,
|
| 2779 |
+
"grad_norm": 26.271549224853516,
|
| 2780 |
+
"learning_rate": 3.8873748016269305e-07,
|
| 2781 |
+
"loss": 14.222109985351562,
|
| 2782 |
+
"step": 7900
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 2.486766048068147,
|
| 2786 |
+
"grad_norm": 23.29692268371582,
|
| 2787 |
+
"learning_rate": 3.5643452801635555e-07,
|
| 2788 |
+
"loss": 11.57226104736328,
|
| 2789 |
+
"step": 7920
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 2.4930469003012847,
|
| 2793 |
+
"grad_norm": 27.596288681030273,
|
| 2794 |
+
"learning_rate": 3.255229403414245e-07,
|
| 2795 |
+
"loss": 14.638540649414063,
|
| 2796 |
+
"step": 7940
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 2.499327752534422,
|
| 2800 |
+
"grad_norm": 24.63420295715332,
|
| 2801 |
+
"learning_rate": 2.960044623920549e-07,
|
| 2802 |
+
"loss": 12.4768310546875,
|
| 2803 |
+
"step": 7960
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 2.505608604767559,
|
| 2807 |
+
"grad_norm": 22.2781982421875,
|
| 2808 |
+
"learning_rate": 2.6788076076804945e-07,
|
| 2809 |
+
"loss": 12.002202606201172,
|
| 2810 |
+
"step": 7980
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 2.511889457000697,
|
| 2814 |
+
"grad_norm": 27.58782196044922,
|
| 2815 |
+
"learning_rate": 2.4115342332078074e-07,
|
| 2816 |
+
"loss": 12.773026275634766,
|
| 2817 |
+
"step": 8000
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 2.518170309233834,
|
| 2821 |
+
"grad_norm": 24.276607513427734,
|
| 2822 |
+
"learning_rate": 2.1582395906353525e-07,
|
| 2823 |
+
"loss": 14.970751953125,
|
| 2824 |
+
"step": 8020
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 2.5244511614669713,
|
| 2828 |
+
"grad_norm": 20.134780883789062,
|
| 2829 |
+
"learning_rate": 1.9189379808632056e-07,
|
| 2830 |
+
"loss": 12.140039825439453,
|
| 2831 |
+
"step": 8040
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 2.530732013700109,
|
| 2835 |
+
"grad_norm": 22.902650833129883,
|
| 2836 |
+
"learning_rate": 1.6936429147511312e-07,
|
| 2837 |
+
"loss": 12.26349334716797,
|
| 2838 |
+
"step": 8060
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 2.5370128659332463,
|
| 2842 |
+
"grad_norm": 30.455528259277344,
|
| 2843 |
+
"learning_rate": 1.4823671123558315e-07,
|
| 2844 |
+
"loss": 11.641680908203124,
|
| 2845 |
+
"step": 8080
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 2.5432937181663835,
|
| 2849 |
+
"grad_norm": 14.522154808044434,
|
| 2850 |
+
"learning_rate": 1.2851225022128266e-07,
|
| 2851 |
+
"loss": 13.1942138671875,
|
| 2852 |
+
"step": 8100
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 2.549574570399521,
|
| 2856 |
+
"grad_norm": 23.315832138061523,
|
| 2857 |
+
"learning_rate": 1.1019202206628554e-07,
|
| 2858 |
+
"loss": 12.113414001464843,
|
| 2859 |
+
"step": 8120
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 2.5558554226326584,
|
| 2863 |
+
"grad_norm": 27.41429901123047,
|
| 2864 |
+
"learning_rate": 9.327706112232104e-08,
|
| 2865 |
+
"loss": 12.481129455566407,
|
| 2866 |
+
"step": 8140
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 2.5621362748657956,
|
| 2870 |
+
"grad_norm": 21.282180786132812,
|
| 2871 |
+
"learning_rate": 7.776832240037335e-08,
|
| 2872 |
+
"loss": 12.50493927001953,
|
| 2873 |
+
"step": 8160
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 2.5684171270989333,
|
| 2877 |
+
"grad_norm": 12.488130569458008,
|
| 2878 |
+
"learning_rate": 6.366668151675804e-08,
|
| 2879 |
+
"loss": 11.507315063476563,
|
| 2880 |
+
"step": 8180
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 2.5746979793320706,
|
| 2884 |
+
"grad_norm": 17.7291316986084,
|
| 2885 |
+
"learning_rate": 5.09729346436949e-08,
|
| 2886 |
+
"loss": 11.453237152099609,
|
| 2887 |
+
"step": 8200
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 2.580978831565208,
|
| 2891 |
+
"grad_norm": 36.775089263916016,
|
| 2892 |
+
"learning_rate": 3.968779846434123e-08,
|
| 2893 |
+
"loss": 11.382515716552735,
|
| 2894 |
+
"step": 8220
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 2.5872596837983455,
|
| 2898 |
+
"grad_norm": 12.098844528198242,
|
| 2899 |
+
"learning_rate": 2.9811910132340746e-08,
|
| 2900 |
+
"loss": 11.915441131591797,
|
| 2901 |
+
"step": 8240
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 2.5935405360314827,
|
| 2905 |
+
"grad_norm": 15.425002098083496,
|
| 2906 |
+
"learning_rate": 2.1345827235846884e-08,
|
| 2907 |
+
"loss": 11.09002456665039,
|
| 2908 |
+
"step": 8260
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 2.59982138826462,
|
| 2912 |
+
"grad_norm": 27.429662704467773,
|
| 2913 |
+
"learning_rate": 1.429002776604238e-08,
|
| 2914 |
+
"loss": 12.257697296142577,
|
| 2915 |
+
"step": 8280
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 2.6061022404977576,
|
| 2919 |
+
"grad_norm": 18.51946449279785,
|
| 2920 |
+
"learning_rate": 8.644910090147007e-09,
|
| 2921 |
+
"loss": 12.654732513427735,
|
| 2922 |
+
"step": 8300
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 2.612383092730895,
|
| 2926 |
+
"grad_norm": 15.994540214538574,
|
| 2927 |
+
"learning_rate": 4.588983511472078e-09,
|
| 2928 |
+
"loss": 11.03597640991211,
|
| 2929 |
+
"step": 8320
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 2.618663944964032,
|
| 2933 |
+
"grad_norm": 26.712631225585938,
|
| 2934 |
+
"learning_rate": 1.695539495108478e-09,
|
| 2935 |
+
"loss": 12.659199523925782,
|
| 2936 |
+
"step": 8340
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 2.6249447971971698,
|
| 2940 |
+
"grad_norm": 29.179128646850586,
|
| 2941 |
+
"learning_rate": 2.1348835171630666e-10,
|
| 2942 |
+
"loss": 14.368611145019532,
|
| 2943 |
+
"step": 8360
|
| 2944 |
+
}
|
| 2945 |
+
],
|
| 2946 |
+
"logging_steps": 20,
|
| 2947 |
+
"max_steps": 8362,
|
| 2948 |
+
"num_input_tokens_seen": 0,
|
| 2949 |
+
"num_train_epochs": 3,
|
| 2950 |
+
"save_steps": 1000000000,
|
| 2951 |
+
"stateful_callbacks": {
|
| 2952 |
+
"TrainerControl": {
|
| 2953 |
+
"args": {
|
| 2954 |
+
"should_epoch_stop": false,
|
| 2955 |
+
"should_evaluate": false,
|
| 2956 |
+
"should_log": false,
|
| 2957 |
+
"should_save": true,
|
| 2958 |
+
"should_training_stop": true
|
| 2959 |
+
},
|
| 2960 |
+
"attributes": {}
|
| 2961 |
+
}
|
| 2962 |
+
},
|
| 2963 |
+
"total_flos": 2.642380317764352e+16,
|
| 2964 |
+
"train_batch_size": 1,
|
| 2965 |
+
"trial_name": null,
|
| 2966 |
+
"trial_params": null
|
| 2967 |
+
}
|