Upload folder using huggingface_hub
Browse files
slots/12/checkpoint-6232/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/12/checkpoint-6232/trainer_state.json
ADDED
|
@@ -0,0 +1,2218 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9568878377184804,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 6232,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0003140426116568692,
|
| 14 |
+
"grad_norm": 13.222649574279785,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 18.71498680114746,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0062808522331373835,
|
| 21 |
+
"grad_norm": 8.995875358581543,
|
| 22 |
+
"learning_rate": 4.999942834660501e-05,
|
| 23 |
+
"loss": 14.314599288137336,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.012561704466274767,
|
| 28 |
+
"grad_norm": 19.639244079589844,
|
| 29 |
+
"learning_rate": 4.999758461765953e-05,
|
| 30 |
+
"loss": 12.460567474365234,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.018842556699412153,
|
| 35 |
+
"grad_norm": 35.52855682373047,
|
| 36 |
+
"learning_rate": 4.999446714274127e-05,
|
| 37 |
+
"loss": 11.741751098632813,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.025123408932549534,
|
| 42 |
+
"grad_norm": 49.0975456237793,
|
| 43 |
+
"learning_rate": 4.999007608070043e-05,
|
| 44 |
+
"loss": 10.829270172119141,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03140426116568692,
|
| 49 |
+
"grad_norm": 22.369361877441406,
|
| 50 |
+
"learning_rate": 4.9984078259584456e-05,
|
| 51 |
+
"loss": 8.55879898071289,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.037685113398824306,
|
| 56 |
+
"grad_norm": 31.618133544921875,
|
| 57 |
+
"learning_rate": 4.997666984880032e-05,
|
| 58 |
+
"loss": 8.94452133178711,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.04396596563196169,
|
| 63 |
+
"grad_norm": 38.25788497924805,
|
| 64 |
+
"learning_rate": 4.9968325679939295e-05,
|
| 65 |
+
"loss": 9.447367858886718,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.05024681786509907,
|
| 70 |
+
"grad_norm": 45.11667251586914,
|
| 71 |
+
"learning_rate": 4.995816789488419e-05,
|
| 72 |
+
"loss": 7.190391540527344,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.056527670098236456,
|
| 77 |
+
"grad_norm": 31.343303680419922,
|
| 78 |
+
"learning_rate": 4.994660098304831e-05,
|
| 79 |
+
"loss": 7.509940338134766,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06280852233137384,
|
| 84 |
+
"grad_norm": 29.660924911499023,
|
| 85 |
+
"learning_rate": 4.993362559749423e-05,
|
| 86 |
+
"loss": 6.391555404663086,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06908937456451122,
|
| 91 |
+
"grad_norm": 26.095754623413086,
|
| 92 |
+
"learning_rate": 4.9919242470806285e-05,
|
| 93 |
+
"loss": 5.232915878295898,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.07537022679764861,
|
| 98 |
+
"grad_norm": 23.29080581665039,
|
| 99 |
+
"learning_rate": 4.990345241504926e-05,
|
| 100 |
+
"loss": 5.611338043212891,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.08165107903078599,
|
| 105 |
+
"grad_norm": 39.18562316894531,
|
| 106 |
+
"learning_rate": 4.98862563217225e-05,
|
| 107 |
+
"loss": 4.7869010925292965,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.08793193126392337,
|
| 112 |
+
"grad_norm": 26.235456466674805,
|
| 113 |
+
"learning_rate": 4.9867655161709595e-05,
|
| 114 |
+
"loss": 3.8826725006103517,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.09421278349706076,
|
| 119 |
+
"grad_norm": 21.395919799804688,
|
| 120 |
+
"learning_rate": 4.984764998522357e-05,
|
| 121 |
+
"loss": 4.216790008544922,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.10049363573019814,
|
| 126 |
+
"grad_norm": 29.157358169555664,
|
| 127 |
+
"learning_rate": 4.9826241921747555e-05,
|
| 128 |
+
"loss": 2.879365921020508,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.10677448796333552,
|
| 133 |
+
"grad_norm": 12.862876892089844,
|
| 134 |
+
"learning_rate": 4.9803432179971055e-05,
|
| 135 |
+
"loss": 3.0516056060791015,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.11305534019647291,
|
| 140 |
+
"grad_norm": 12.190792083740234,
|
| 141 |
+
"learning_rate": 4.9779222047721693e-05,
|
| 142 |
+
"loss": 2.235590934753418,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1193361924296103,
|
| 147 |
+
"grad_norm": 6.71710729598999,
|
| 148 |
+
"learning_rate": 4.9753612891892514e-05,
|
| 149 |
+
"loss": 2.179483985900879,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.12561704466274767,
|
| 154 |
+
"grad_norm": 3.582365036010742,
|
| 155 |
+
"learning_rate": 4.9726606158364765e-05,
|
| 156 |
+
"loss": 1.6689449310302735,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.13189789689588505,
|
| 161 |
+
"grad_norm": 7.28436803817749,
|
| 162 |
+
"learning_rate": 4.9698203371926305e-05,
|
| 163 |
+
"loss": 1.5380577087402343,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.13817874912902245,
|
| 168 |
+
"grad_norm": 7.6366353034973145,
|
| 169 |
+
"learning_rate": 4.966840613618552e-05,
|
| 170 |
+
"loss": 1.4540648460388184,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.14445960136215982,
|
| 175 |
+
"grad_norm": 8.194938659667969,
|
| 176 |
+
"learning_rate": 4.963721613348074e-05,
|
| 177 |
+
"loss": 1.1600905418395997,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.15074045359529722,
|
| 182 |
+
"grad_norm": 2.6455771923065186,
|
| 183 |
+
"learning_rate": 4.960463512478531e-05,
|
| 184 |
+
"loss": 1.041440486907959,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1570213058284346,
|
| 189 |
+
"grad_norm": 4.079453945159912,
|
| 190 |
+
"learning_rate": 4.957066494960813e-05,
|
| 191 |
+
"loss": 0.9727105140686035,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.16330215806157197,
|
| 196 |
+
"grad_norm": 4.176381587982178,
|
| 197 |
+
"learning_rate": 4.953530752588979e-05,
|
| 198 |
+
"loss": 0.7286320209503174,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.16958301029470937,
|
| 203 |
+
"grad_norm": 3.183208703994751,
|
| 204 |
+
"learning_rate": 4.949856484989432e-05,
|
| 205 |
+
"loss": 0.8421343803405762,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.17586386252784675,
|
| 210 |
+
"grad_norm": 2.9324145317077637,
|
| 211 |
+
"learning_rate": 4.946043899609646e-05,
|
| 212 |
+
"loss": 0.8696622848510742,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.18214471476098412,
|
| 217 |
+
"grad_norm": 1.9436066150665283,
|
| 218 |
+
"learning_rate": 4.9420932117064535e-05,
|
| 219 |
+
"loss": 0.632416820526123,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.18842556699412152,
|
| 224 |
+
"grad_norm": 2.4007198810577393,
|
| 225 |
+
"learning_rate": 4.9380046443338905e-05,
|
| 226 |
+
"loss": 0.5935386657714844,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.1947064192272589,
|
| 231 |
+
"grad_norm": 2.132748603820801,
|
| 232 |
+
"learning_rate": 4.933778428330609e-05,
|
| 233 |
+
"loss": 0.543156623840332,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.20098727146039627,
|
| 238 |
+
"grad_norm": 2.3025615215301514,
|
| 239 |
+
"learning_rate": 4.929414802306837e-05,
|
| 240 |
+
"loss": 0.5939158916473388,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.20726812369353367,
|
| 245 |
+
"grad_norm": 0.910429060459137,
|
| 246 |
+
"learning_rate": 4.9249140126309104e-05,
|
| 247 |
+
"loss": 0.49994521141052245,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.21354897592667105,
|
| 252 |
+
"grad_norm": 1.484687328338623,
|
| 253 |
+
"learning_rate": 4.9202763134153615e-05,
|
| 254 |
+
"loss": 0.5400276184082031,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.21982982815980842,
|
| 259 |
+
"grad_norm": 1.6964185237884521,
|
| 260 |
+
"learning_rate": 4.915501966502576e-05,
|
| 261 |
+
"loss": 0.44618911743164064,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.22611068039294582,
|
| 266 |
+
"grad_norm": 1.9460206031799316,
|
| 267 |
+
"learning_rate": 4.910591241450002e-05,
|
| 268 |
+
"loss": 0.4275932788848877,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.2323915326260832,
|
| 273 |
+
"grad_norm": 1.8712719678878784,
|
| 274 |
+
"learning_rate": 4.905544415514936e-05,
|
| 275 |
+
"loss": 0.4287854194641113,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2386723848592206,
|
| 280 |
+
"grad_norm": 1.1180524826049805,
|
| 281 |
+
"learning_rate": 4.900361773638873e-05,
|
| 282 |
+
"loss": 0.3843901872634888,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.24495323709235797,
|
| 287 |
+
"grad_norm": 0.9346688389778137,
|
| 288 |
+
"learning_rate": 4.895043608431408e-05,
|
| 289 |
+
"loss": 0.45698018074035646,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.25123408932549535,
|
| 294 |
+
"grad_norm": 2.003167152404785,
|
| 295 |
+
"learning_rate": 4.8895902201537245e-05,
|
| 296 |
+
"loss": 0.3375147581100464,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.25751494155863275,
|
| 301 |
+
"grad_norm": 1.505427360534668,
|
| 302 |
+
"learning_rate": 4.884001916701639e-05,
|
| 303 |
+
"loss": 0.35217485427856443,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2637957937917701,
|
| 308 |
+
"grad_norm": 1.0035185813903809,
|
| 309 |
+
"learning_rate": 4.878279013588215e-05,
|
| 310 |
+
"loss": 0.3541921377182007,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2700766460249075,
|
| 315 |
+
"grad_norm": 0.7123732566833496,
|
| 316 |
+
"learning_rate": 4.872421833925955e-05,
|
| 317 |
+
"loss": 0.40600271224975587,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.2763574982580449,
|
| 322 |
+
"grad_norm": 1.2544392347335815,
|
| 323 |
+
"learning_rate": 4.8664307084085516e-05,
|
| 324 |
+
"loss": 0.26212928295135496,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2826383504911823,
|
| 329 |
+
"grad_norm": 1.4210506677627563,
|
| 330 |
+
"learning_rate": 4.860305975292221e-05,
|
| 331 |
+
"loss": 0.3331881523132324,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.28891920272431965,
|
| 336 |
+
"grad_norm": 1.0728765726089478,
|
| 337 |
+
"learning_rate": 4.854047980376601e-05,
|
| 338 |
+
"loss": 0.29691243171691895,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.29520005495745705,
|
| 343 |
+
"grad_norm": 0.7383350729942322,
|
| 344 |
+
"learning_rate": 4.847657076985233e-05,
|
| 345 |
+
"loss": 0.2863906383514404,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.30148090719059445,
|
| 350 |
+
"grad_norm": 1.553955316543579,
|
| 351 |
+
"learning_rate": 4.8411336259456085e-05,
|
| 352 |
+
"loss": 0.30426201820373533,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3077617594237318,
|
| 357 |
+
"grad_norm": 0.6390340924263,
|
| 358 |
+
"learning_rate": 4.834477995568798e-05,
|
| 359 |
+
"loss": 0.2549339056015015,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3140426116568692,
|
| 364 |
+
"grad_norm": 1.8312214612960815,
|
| 365 |
+
"learning_rate": 4.8276905616286596e-05,
|
| 366 |
+
"loss": 0.34730048179626466,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3203234638900066,
|
| 371 |
+
"grad_norm": 1.6143252849578857,
|
| 372 |
+
"learning_rate": 4.820771707340619e-05,
|
| 373 |
+
"loss": 0.2739202260971069,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.32660431612314395,
|
| 378 |
+
"grad_norm": 0.7645473480224609,
|
| 379 |
+
"learning_rate": 4.813721823340034e-05,
|
| 380 |
+
"loss": 0.2440113306045532,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.33288516835628135,
|
| 385 |
+
"grad_norm": 1.8165767192840576,
|
| 386 |
+
"learning_rate": 4.8065413076601416e-05,
|
| 387 |
+
"loss": 0.1889214277267456,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.33916602058941875,
|
| 392 |
+
"grad_norm": 0.9012591242790222,
|
| 393 |
+
"learning_rate": 4.799230565709584e-05,
|
| 394 |
+
"loss": 0.20236475467681886,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3454468728225561,
|
| 399 |
+
"grad_norm": 0.9148083925247192,
|
| 400 |
+
"learning_rate": 4.7917900102495185e-05,
|
| 401 |
+
"loss": 0.24533369541168212,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.3517277250556935,
|
| 406 |
+
"grad_norm": 1.9009038209915161,
|
| 407 |
+
"learning_rate": 4.784220061370314e-05,
|
| 408 |
+
"loss": 0.19988182783126832,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3580085772888309,
|
| 413 |
+
"grad_norm": 0.6809736490249634,
|
| 414 |
+
"learning_rate": 4.776521146467833e-05,
|
| 415 |
+
"loss": 0.23843612670898437,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.36428942952196824,
|
| 420 |
+
"grad_norm": 0.2764051854610443,
|
| 421 |
+
"learning_rate": 4.768693700219301e-05,
|
| 422 |
+
"loss": 0.25783936977386473,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.37057028175510565,
|
| 427 |
+
"grad_norm": 2.5077197551727295,
|
| 428 |
+
"learning_rate": 4.7607381645587644e-05,
|
| 429 |
+
"loss": 0.24673666954040527,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.37685113398824305,
|
| 434 |
+
"grad_norm": 1.353865146636963,
|
| 435 |
+
"learning_rate": 4.75265498865214e-05,
|
| 436 |
+
"loss": 0.21446623802185058,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3831319862213804,
|
| 441 |
+
"grad_norm": 0.7787877917289734,
|
| 442 |
+
"learning_rate": 4.744444628871855e-05,
|
| 443 |
+
"loss": 0.24493517875671386,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.3894128384545178,
|
| 448 |
+
"grad_norm": 0.4324120283126831,
|
| 449 |
+
"learning_rate": 4.736107548771077e-05,
|
| 450 |
+
"loss": 0.2510174989700317,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3956936906876552,
|
| 455 |
+
"grad_norm": 0.9496787786483765,
|
| 456 |
+
"learning_rate": 4.72764421905755e-05,
|
| 457 |
+
"loss": 0.2243569612503052,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.40197454292079254,
|
| 462 |
+
"grad_norm": 0.36113232374191284,
|
| 463 |
+
"learning_rate": 4.719055117567011e-05,
|
| 464 |
+
"loss": 0.21692416667938233,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.40825539515392995,
|
| 469 |
+
"grad_norm": 0.989080548286438,
|
| 470 |
+
"learning_rate": 4.710340729236212e-05,
|
| 471 |
+
"loss": 0.17741645574569703,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.41453624738706735,
|
| 476 |
+
"grad_norm": 0.798384428024292,
|
| 477 |
+
"learning_rate": 4.701501546075548e-05,
|
| 478 |
+
"loss": 0.183823299407959,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.4208170996202047,
|
| 483 |
+
"grad_norm": 0.5087212920188904,
|
| 484 |
+
"learning_rate": 4.6925380671412695e-05,
|
| 485 |
+
"loss": 0.24055900573730468,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.4270979518533421,
|
| 490 |
+
"grad_norm": 0.5015459060668945,
|
| 491 |
+
"learning_rate": 4.68345079850731e-05,
|
| 492 |
+
"loss": 0.20093884468078613,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.4333788040864795,
|
| 497 |
+
"grad_norm": 2.0808253288269043,
|
| 498 |
+
"learning_rate": 4.674240253236714e-05,
|
| 499 |
+
"loss": 0.2449209451675415,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.43965965631961684,
|
| 504 |
+
"grad_norm": 0.9437074661254883,
|
| 505 |
+
"learning_rate": 4.664906951352669e-05,
|
| 506 |
+
"loss": 0.1986350893974304,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.44594050855275424,
|
| 511 |
+
"grad_norm": 0.41122740507125854,
|
| 512 |
+
"learning_rate": 4.655451419809142e-05,
|
| 513 |
+
"loss": 0.2406865119934082,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.45222136078589165,
|
| 518 |
+
"grad_norm": 0.3834715187549591,
|
| 519 |
+
"learning_rate": 4.645874192461137e-05,
|
| 520 |
+
"loss": 0.24607343673706056,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.458502213019029,
|
| 525 |
+
"grad_norm": 10.456542015075684,
|
| 526 |
+
"learning_rate": 4.6361758100345404e-05,
|
| 527 |
+
"loss": 0.20838961601257325,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.4647830652521664,
|
| 532 |
+
"grad_norm": 0.20652760565280914,
|
| 533 |
+
"learning_rate": 4.6263568200956034e-05,
|
| 534 |
+
"loss": 0.17515504360198975,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.4710639174853038,
|
| 539 |
+
"grad_norm": 0.833929181098938,
|
| 540 |
+
"learning_rate": 4.6164177770200226e-05,
|
| 541 |
+
"loss": 0.16210874319076538,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.4773447697184412,
|
| 546 |
+
"grad_norm": 0.6646195650100708,
|
| 547 |
+
"learning_rate": 4.606359241961637e-05,
|
| 548 |
+
"loss": 0.17271146774291993,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.48362562195157854,
|
| 553 |
+
"grad_norm": 0.45452389121055603,
|
| 554 |
+
"learning_rate": 4.596181782820751e-05,
|
| 555 |
+
"loss": 0.12133709192276002,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.48990647418471595,
|
| 560 |
+
"grad_norm": 0.2559211254119873,
|
| 561 |
+
"learning_rate": 4.585885974212068e-05,
|
| 562 |
+
"loss": 0.19305528402328492,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.49618732641785335,
|
| 567 |
+
"grad_norm": 0.3878309726715088,
|
| 568 |
+
"learning_rate": 4.575472397432248e-05,
|
| 569 |
+
"loss": 0.19771596193313598,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.5024681786509907,
|
| 574 |
+
"grad_norm": 0.5782968997955322,
|
| 575 |
+
"learning_rate": 4.564941640427088e-05,
|
| 576 |
+
"loss": 0.1698690176010132,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.508749030884128,
|
| 581 |
+
"grad_norm": 0.4471549093723297,
|
| 582 |
+
"learning_rate": 4.5542942977583285e-05,
|
| 583 |
+
"loss": 0.15614293813705443,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.5150298831172655,
|
| 588 |
+
"grad_norm": 0.5035043954849243,
|
| 589 |
+
"learning_rate": 4.543530970570081e-05,
|
| 590 |
+
"loss": 0.13844941854476928,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.5213107353504028,
|
| 595 |
+
"grad_norm": 0.4261391758918762,
|
| 596 |
+
"learning_rate": 4.532652266554893e-05,
|
| 597 |
+
"loss": 0.17091692686080934,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.5275915875835402,
|
| 602 |
+
"grad_norm": 1.0388013124465942,
|
| 603 |
+
"learning_rate": 4.521658799919431e-05,
|
| 604 |
+
"loss": 0.16304466724395753,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.5338724398166776,
|
| 609 |
+
"grad_norm": 0.3387233316898346,
|
| 610 |
+
"learning_rate": 4.510551191349813e-05,
|
| 611 |
+
"loss": 0.14147281646728516,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.540153292049815,
|
| 616 |
+
"grad_norm": 0.47025224566459656,
|
| 617 |
+
"learning_rate": 4.499330067976552e-05,
|
| 618 |
+
"loss": 0.14872909784317018,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.5464341442829523,
|
| 623 |
+
"grad_norm": 0.4901464879512787,
|
| 624 |
+
"learning_rate": 4.487996063339158e-05,
|
| 625 |
+
"loss": 0.13481216430664061,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.5527149965160898,
|
| 630 |
+
"grad_norm": 1.1053012609481812,
|
| 631 |
+
"learning_rate": 4.476549817350367e-05,
|
| 632 |
+
"loss": 0.11047512292861938,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5589958487492271,
|
| 637 |
+
"grad_norm": 0.39822623133659363,
|
| 638 |
+
"learning_rate": 4.464991976260009e-05,
|
| 639 |
+
"loss": 0.2455967903137207,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.5652767009823646,
|
| 644 |
+
"grad_norm": 0.33123093843460083,
|
| 645 |
+
"learning_rate": 4.453323192618521e-05,
|
| 646 |
+
"loss": 0.35040247440338135,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.571557553215502,
|
| 651 |
+
"grad_norm": 0.3363930284976959,
|
| 652 |
+
"learning_rate": 4.4415441252401094e-05,
|
| 653 |
+
"loss": 0.09571367502212524,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5778384054486393,
|
| 658 |
+
"grad_norm": 1.574691653251648,
|
| 659 |
+
"learning_rate": 4.4296554391655464e-05,
|
| 660 |
+
"loss": 0.11115046739578247,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5841192576817767,
|
| 665 |
+
"grad_norm": 2.8482120037078857,
|
| 666 |
+
"learning_rate": 4.417657805624629e-05,
|
| 667 |
+
"loss": 0.17036218643188478,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5904001099149141,
|
| 672 |
+
"grad_norm": 0.21437305212020874,
|
| 673 |
+
"learning_rate": 4.405551901998277e-05,
|
| 674 |
+
"loss": 0.1317328095436096,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.5966809621480514,
|
| 679 |
+
"grad_norm": 0.5809491872787476,
|
| 680 |
+
"learning_rate": 4.39333841178029e-05,
|
| 681 |
+
"loss": 0.11890696287155152,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.6029618143811889,
|
| 686 |
+
"grad_norm": 0.465741366147995,
|
| 687 |
+
"learning_rate": 4.3810180245387574e-05,
|
| 688 |
+
"loss": 0.08567101359367371,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6092426666143262,
|
| 693 |
+
"grad_norm": 0.29012948274612427,
|
| 694 |
+
"learning_rate": 4.368591435877125e-05,
|
| 695 |
+
"loss": 0.14835361242294312,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.6155235188474636,
|
| 700 |
+
"grad_norm": 0.3374810516834259,
|
| 701 |
+
"learning_rate": 4.3560593473949254e-05,
|
| 702 |
+
"loss": 0.1041601061820984,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.621804371080601,
|
| 707 |
+
"grad_norm": 0.21143510937690735,
|
| 708 |
+
"learning_rate": 4.343422466648159e-05,
|
| 709 |
+
"loss": 0.08853803277015686,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.6280852233137384,
|
| 714 |
+
"grad_norm": 0.8877353668212891,
|
| 715 |
+
"learning_rate": 4.330681507109352e-05,
|
| 716 |
+
"loss": 0.11505429744720459,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.6343660755468757,
|
| 721 |
+
"grad_norm": 0.8372551798820496,
|
| 722 |
+
"learning_rate": 4.3178371881272716e-05,
|
| 723 |
+
"loss": 0.15943719148635865,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.6406469277800132,
|
| 728 |
+
"grad_norm": 0.3456370532512665,
|
| 729 |
+
"learning_rate": 4.3048902348863116e-05,
|
| 730 |
+
"loss": 0.15141594409942627,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.6469277800131505,
|
| 735 |
+
"grad_norm": 0.5323220491409302,
|
| 736 |
+
"learning_rate": 4.2918413783655484e-05,
|
| 737 |
+
"loss": 0.09476671218872071,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6532086322462879,
|
| 742 |
+
"grad_norm": 3.5620996952056885,
|
| 743 |
+
"learning_rate": 4.2786913552974744e-05,
|
| 744 |
+
"loss": 0.22961113452911378,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.6594894844794253,
|
| 749 |
+
"grad_norm": 3.3487048149108887,
|
| 750 |
+
"learning_rate": 4.265440908126395e-05,
|
| 751 |
+
"loss": 0.1464246392250061,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6657703367125627,
|
| 756 |
+
"grad_norm": 0.358454167842865,
|
| 757 |
+
"learning_rate": 4.252090784966517e-05,
|
| 758 |
+
"loss": 0.17966300249099731,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6720511889457,
|
| 763 |
+
"grad_norm": 0.6528313755989075,
|
| 764 |
+
"learning_rate": 4.238641739559708e-05,
|
| 765 |
+
"loss": 0.15987756252288818,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6783320411788375,
|
| 770 |
+
"grad_norm": 0.3723723292350769,
|
| 771 |
+
"learning_rate": 4.22509453123294e-05,
|
| 772 |
+
"loss": 0.17160117626190186,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.6846128934119748,
|
| 777 |
+
"grad_norm": 0.5360041856765747,
|
| 778 |
+
"learning_rate": 4.211449924855418e-05,
|
| 779 |
+
"loss": 0.08139219880104065,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.6908937456451122,
|
| 784 |
+
"grad_norm": 0.7036887407302856,
|
| 785 |
+
"learning_rate": 4.1977086907953986e-05,
|
| 786 |
+
"loss": 0.1261943459510803,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.6971745978782496,
|
| 791 |
+
"grad_norm": 0.6446220278739929,
|
| 792 |
+
"learning_rate": 4.183871604876691e-05,
|
| 793 |
+
"loss": 0.12319077253341675,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.703455450111387,
|
| 798 |
+
"grad_norm": 0.8421915769577026,
|
| 799 |
+
"learning_rate": 4.169939448334859e-05,
|
| 800 |
+
"loss": 0.13604164123535156,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.7097363023445243,
|
| 805 |
+
"grad_norm": 0.0917246863245964,
|
| 806 |
+
"learning_rate": 4.155913007773109e-05,
|
| 807 |
+
"loss": 0.11214659214019776,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.7160171545776618,
|
| 812 |
+
"grad_norm": 2.0724904537200928,
|
| 813 |
+
"learning_rate": 4.141793075117879e-05,
|
| 814 |
+
"loss": 0.15922495126724243,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.7222980068107991,
|
| 819 |
+
"grad_norm": 0.33200934529304504,
|
| 820 |
+
"learning_rate": 4.127580447574131e-05,
|
| 821 |
+
"loss": 0.11984597444534302,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.7285788590439365,
|
| 826 |
+
"grad_norm": 0.23812741041183472,
|
| 827 |
+
"learning_rate": 4.113275927580337e-05,
|
| 828 |
+
"loss": 0.11355631351470948,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.734859711277074,
|
| 833 |
+
"grad_norm": 1.4909952878952026,
|
| 834 |
+
"learning_rate": 4.098880322763172e-05,
|
| 835 |
+
"loss": 0.11960653066635132,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.7411405635102113,
|
| 840 |
+
"grad_norm": 0.7009281516075134,
|
| 841 |
+
"learning_rate": 4.0843944458919216e-05,
|
| 842 |
+
"loss": 0.10994844436645508,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.7474214157433486,
|
| 847 |
+
"grad_norm": 1.218519926071167,
|
| 848 |
+
"learning_rate": 4.069819114832591e-05,
|
| 849 |
+
"loss": 0.10543915033340454,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.7537022679764861,
|
| 854 |
+
"grad_norm": 0.5172832012176514,
|
| 855 |
+
"learning_rate": 4.0551551525017264e-05,
|
| 856 |
+
"loss": 0.09124865531921386,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.7599831202096234,
|
| 861 |
+
"grad_norm": 0.7078964114189148,
|
| 862 |
+
"learning_rate": 4.0404033868199554e-05,
|
| 863 |
+
"loss": 0.10676212310791015,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.7662639724427608,
|
| 868 |
+
"grad_norm": 0.35599201917648315,
|
| 869 |
+
"learning_rate": 4.0255646506652426e-05,
|
| 870 |
+
"loss": 0.07489743828773499,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.7725448246758982,
|
| 875 |
+
"grad_norm": 0.9615290760993958,
|
| 876 |
+
"learning_rate": 4.010639781825867e-05,
|
| 877 |
+
"loss": 0.10462805032730102,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.7788256769090356,
|
| 882 |
+
"grad_norm": 0.3290087878704071,
|
| 883 |
+
"learning_rate": 3.9956296229531175e-05,
|
| 884 |
+
"loss": 0.10868211984634399,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.7851065291421729,
|
| 889 |
+
"grad_norm": 0.5029305219650269,
|
| 890 |
+
"learning_rate": 3.980535021513723e-05,
|
| 891 |
+
"loss": 0.06722604036331177,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.7913873813753104,
|
| 896 |
+
"grad_norm": 0.3802237808704376,
|
| 897 |
+
"learning_rate": 3.9653568297419975e-05,
|
| 898 |
+
"loss": 0.10265893936157226,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.7976682336084477,
|
| 903 |
+
"grad_norm": 1.5072736740112305,
|
| 904 |
+
"learning_rate": 3.950095904591729e-05,
|
| 905 |
+
"loss": 0.07902958393096923,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.8039490858415851,
|
| 910 |
+
"grad_norm": 1.0115009546279907,
|
| 911 |
+
"learning_rate": 3.9347531076877944e-05,
|
| 912 |
+
"loss": 0.08553307056427002,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.8102299380747225,
|
| 917 |
+
"grad_norm": 0.1543072760105133,
|
| 918 |
+
"learning_rate": 3.919329305277511e-05,
|
| 919 |
+
"loss": 0.11212378740310669,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.8165107903078599,
|
| 924 |
+
"grad_norm": 0.2683345377445221,
|
| 925 |
+
"learning_rate": 3.9038253681817314e-05,
|
| 926 |
+
"loss": 0.08978410959243774,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.8227916425409972,
|
| 931 |
+
"grad_norm": 0.8963088393211365,
|
| 932 |
+
"learning_rate": 3.888242171745677e-05,
|
| 933 |
+
"loss": 0.079929518699646,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.8290724947741347,
|
| 938 |
+
"grad_norm": 0.2084486186504364,
|
| 939 |
+
"learning_rate": 3.872580595789513e-05,
|
| 940 |
+
"loss": 0.1274382472038269,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.835353347007272,
|
| 945 |
+
"grad_norm": 0.6193188428878784,
|
| 946 |
+
"learning_rate": 3.856841524558678e-05,
|
| 947 |
+
"loss": 0.15513088703155517,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.8416341992404094,
|
| 952 |
+
"grad_norm": 0.2659473419189453,
|
| 953 |
+
"learning_rate": 3.841025846673957e-05,
|
| 954 |
+
"loss": 0.08328367471694946,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.8479150514735468,
|
| 959 |
+
"grad_norm": 0.33367040753364563,
|
| 960 |
+
"learning_rate": 3.8251344550813136e-05,
|
| 961 |
+
"loss": 0.08085548877716064,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.8541959037066842,
|
| 966 |
+
"grad_norm": 0.577295184135437,
|
| 967 |
+
"learning_rate": 3.809168247001473e-05,
|
| 968 |
+
"loss": 0.0950459897518158,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.8604767559398215,
|
| 973 |
+
"grad_norm": 0.41376471519470215,
|
| 974 |
+
"learning_rate": 3.793128123879265e-05,
|
| 975 |
+
"loss": 0.059304440021514894,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.866757608172959,
|
| 980 |
+
"grad_norm": 0.8415532112121582,
|
| 981 |
+
"learning_rate": 3.7770149913327255e-05,
|
| 982 |
+
"loss": 0.09070171713829041,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.8730384604060963,
|
| 987 |
+
"grad_norm": 0.8845179677009583,
|
| 988 |
+
"learning_rate": 3.760829759101974e-05,
|
| 989 |
+
"loss": 0.09670570492744446,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.8793193126392337,
|
| 994 |
+
"grad_norm": 1.0329339504241943,
|
| 995 |
+
"learning_rate": 3.744573340997843e-05,
|
| 996 |
+
"loss": 0.12236303091049194,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.8856001648723711,
|
| 1001 |
+
"grad_norm": 0.3550909459590912,
|
| 1002 |
+
"learning_rate": 3.728246654850287e-05,
|
| 1003 |
+
"loss": 0.11134687662124634,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.8918810171055085,
|
| 1008 |
+
"grad_norm": 0.1834748536348343,
|
| 1009 |
+
"learning_rate": 3.711850622456562e-05,
|
| 1010 |
+
"loss": 0.10767635107040405,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.8981618693386458,
|
| 1015 |
+
"grad_norm": 0.9229864478111267,
|
| 1016 |
+
"learning_rate": 3.695386169529184e-05,
|
| 1017 |
+
"loss": 0.09790213704109192,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.9044427215717833,
|
| 1022 |
+
"grad_norm": 0.797934889793396,
|
| 1023 |
+
"learning_rate": 3.6788542256436565e-05,
|
| 1024 |
+
"loss": 0.09137773513793945,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.9107235738049206,
|
| 1029 |
+
"grad_norm": 0.44148725271224976,
|
| 1030 |
+
"learning_rate": 3.662255724185996e-05,
|
| 1031 |
+
"loss": 0.06808314323425294,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.917004426038058,
|
| 1036 |
+
"grad_norm": 0.4025018513202667,
|
| 1037 |
+
"learning_rate": 3.645591602300026e-05,
|
| 1038 |
+
"loss": 0.0749214231967926,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.9232852782711954,
|
| 1043 |
+
"grad_norm": 0.5347099900245667,
|
| 1044 |
+
"learning_rate": 3.628862800834472e-05,
|
| 1045 |
+
"loss": 0.09171435832977295,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.9295661305043328,
|
| 1050 |
+
"grad_norm": 0.1338285505771637,
|
| 1051 |
+
"learning_rate": 3.6120702642898366e-05,
|
| 1052 |
+
"loss": 0.0905307412147522,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.9358469827374702,
|
| 1057 |
+
"grad_norm": 0.3484551012516022,
|
| 1058 |
+
"learning_rate": 3.59521494076508e-05,
|
| 1059 |
+
"loss": 0.08842827081680298,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.9421278349706076,
|
| 1064 |
+
"grad_norm": 0.14611800014972687,
|
| 1065 |
+
"learning_rate": 3.5782977819040816e-05,
|
| 1066 |
+
"loss": 0.078364896774292,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.9484086872037449,
|
| 1071 |
+
"grad_norm": 1.907283902168274,
|
| 1072 |
+
"learning_rate": 3.561319742841919e-05,
|
| 1073 |
+
"loss": 0.08273177146911621,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.9546895394368824,
|
| 1078 |
+
"grad_norm": 0.30305084586143494,
|
| 1079 |
+
"learning_rate": 3.544281782150936e-05,
|
| 1080 |
+
"loss": 0.09260942339897156,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.9609703916700197,
|
| 1085 |
+
"grad_norm": 0.20987313985824585,
|
| 1086 |
+
"learning_rate": 3.527184861786629e-05,
|
| 1087 |
+
"loss": 0.06498585939407349,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.9672512439031571,
|
| 1092 |
+
"grad_norm": 0.612435519695282,
|
| 1093 |
+
"learning_rate": 3.510029947033323e-05,
|
| 1094 |
+
"loss": 0.0671736478805542,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.9735320961362945,
|
| 1099 |
+
"grad_norm": 0.12430224567651749,
|
| 1100 |
+
"learning_rate": 3.492818006449685e-05,
|
| 1101 |
+
"loss": 0.059837305545806886,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.9798129483694319,
|
| 1106 |
+
"grad_norm": 1.1324347257614136,
|
| 1107 |
+
"learning_rate": 3.4755500118140364e-05,
|
| 1108 |
+
"loss": 0.07871156334877014,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9860938006025692,
|
| 1113 |
+
"grad_norm": 0.2752481698989868,
|
| 1114 |
+
"learning_rate": 3.4582269380694775e-05,
|
| 1115 |
+
"loss": 0.10896846055984497,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.9923746528357067,
|
| 1120 |
+
"grad_norm": 0.5350470542907715,
|
| 1121 |
+
"learning_rate": 3.4408497632688576e-05,
|
| 1122 |
+
"loss": 0.0751093327999115,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.998655505068844,
|
| 1127 |
+
"grad_norm": 1.6723430156707764,
|
| 1128 |
+
"learning_rate": 3.4234194685195405e-05,
|
| 1129 |
+
"loss": 0.07581815123558044,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.004710639174853,
|
| 1134 |
+
"grad_norm": 0.35298222303390503,
|
| 1135 |
+
"learning_rate": 3.4059370379280245e-05,
|
| 1136 |
+
"loss": 0.045219630002975464,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.0109914914079905,
|
| 1141 |
+
"grad_norm": 0.7559930086135864,
|
| 1142 |
+
"learning_rate": 3.388403458544368e-05,
|
| 1143 |
+
"loss": 0.06403674483299256,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.0172723436411277,
|
| 1148 |
+
"grad_norm": 0.21411669254302979,
|
| 1149 |
+
"learning_rate": 3.370819720306469e-05,
|
| 1150 |
+
"loss": 0.03882692754268646,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.0235531958742652,
|
| 1155 |
+
"grad_norm": 0.391485333442688,
|
| 1156 |
+
"learning_rate": 3.3531868159841754e-05,
|
| 1157 |
+
"loss": 0.07030806541442872,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.0298340481074026,
|
| 1162 |
+
"grad_norm": 0.2437785267829895,
|
| 1163 |
+
"learning_rate": 3.335505741123222e-05,
|
| 1164 |
+
"loss": 0.04769740402698517,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.0361149003405399,
|
| 1169 |
+
"grad_norm": 0.15534839034080505,
|
| 1170 |
+
"learning_rate": 3.317777493989037e-05,
|
| 1171 |
+
"loss": 0.0423685759305954,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.0423957525736773,
|
| 1176 |
+
"grad_norm": 0.36984455585479736,
|
| 1177 |
+
"learning_rate": 3.300003075510371e-05,
|
| 1178 |
+
"loss": 0.053946495056152344,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.0486766048068148,
|
| 1183 |
+
"grad_norm": 0.2222166508436203,
|
| 1184 |
+
"learning_rate": 3.282183489222788e-05,
|
| 1185 |
+
"loss": 0.10178368091583252,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.054957457039952,
|
| 1190 |
+
"grad_norm": 0.29375144839286804,
|
| 1191 |
+
"learning_rate": 3.264319741212004e-05,
|
| 1192 |
+
"loss": 0.0629787027835846,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.0612383092730895,
|
| 1197 |
+
"grad_norm": 0.13212697207927704,
|
| 1198 |
+
"learning_rate": 3.246412840057089e-05,
|
| 1199 |
+
"loss": 0.05211652517318725,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.067519161506227,
|
| 1204 |
+
"grad_norm": 0.08366473019123077,
|
| 1205 |
+
"learning_rate": 3.228463796773519e-05,
|
| 1206 |
+
"loss": 0.04285470843315124,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.0738000137393642,
|
| 1211 |
+
"grad_norm": 0.1260441094636917,
|
| 1212 |
+
"learning_rate": 3.2104736247560927e-05,
|
| 1213 |
+
"loss": 0.03178199231624603,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.0800808659725016,
|
| 1218 |
+
"grad_norm": 0.11161878705024719,
|
| 1219 |
+
"learning_rate": 3.192443339721723e-05,
|
| 1220 |
+
"loss": 0.06174847483634949,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.086361718205639,
|
| 1225 |
+
"grad_norm": 0.3836033046245575,
|
| 1226 |
+
"learning_rate": 3.17437395965208e-05,
|
| 1227 |
+
"loss": 0.06333217024803162,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.0926425704387763,
|
| 1232 |
+
"grad_norm": 0.7206199765205383,
|
| 1233 |
+
"learning_rate": 3.1562665047361264e-05,
|
| 1234 |
+
"loss": 0.038184750080108645,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.0989234226719138,
|
| 1239 |
+
"grad_norm": 0.40549972653388977,
|
| 1240 |
+
"learning_rate": 3.1381219973125107e-05,
|
| 1241 |
+
"loss": 0.07620968222618103,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.1052042749050512,
|
| 1246 |
+
"grad_norm": 0.22500963509082794,
|
| 1247 |
+
"learning_rate": 3.11994146181185e-05,
|
| 1248 |
+
"loss": 0.04413818717002869,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.1114851271381885,
|
| 1253 |
+
"grad_norm": 0.07935626059770584,
|
| 1254 |
+
"learning_rate": 3.1017259246988905e-05,
|
| 1255 |
+
"loss": 0.03253397643566132,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.117765979371326,
|
| 1260 |
+
"grad_norm": 1.564283847808838,
|
| 1261 |
+
"learning_rate": 3.0834764144145526e-05,
|
| 1262 |
+
"loss": 0.044573229551315305,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.1240468316044634,
|
| 1267 |
+
"grad_norm": 0.28379613161087036,
|
| 1268 |
+
"learning_rate": 3.0651939613178674e-05,
|
| 1269 |
+
"loss": 0.041404137015342714,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.1303276838376006,
|
| 1274 |
+
"grad_norm": 0.6835076212882996,
|
| 1275 |
+
"learning_rate": 3.046879597627801e-05,
|
| 1276 |
+
"loss": 0.03626242876052856,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.136608536070738,
|
| 1281 |
+
"grad_norm": 0.42547276616096497,
|
| 1282 |
+
"learning_rate": 3.028534357364977e-05,
|
| 1283 |
+
"loss": 0.04724129736423492,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.1428893883038755,
|
| 1288 |
+
"grad_norm": 0.3822947144508362,
|
| 1289 |
+
"learning_rate": 3.010159276293299e-05,
|
| 1290 |
+
"loss": 0.051716703176498416,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.1491702405370128,
|
| 1295 |
+
"grad_norm": 0.11303099244832993,
|
| 1296 |
+
"learning_rate": 2.991755391861465e-05,
|
| 1297 |
+
"loss": 0.048957851529121396,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.1554510927701502,
|
| 1302 |
+
"grad_norm": 0.14571624994277954,
|
| 1303 |
+
"learning_rate": 2.9733237431444017e-05,
|
| 1304 |
+
"loss": 0.042123788595199586,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.1617319450032877,
|
| 1309 |
+
"grad_norm": 0.15363015234470367,
|
| 1310 |
+
"learning_rate": 2.9548653707845903e-05,
|
| 1311 |
+
"loss": 0.037100601196289065,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.168012797236425,
|
| 1316 |
+
"grad_norm": 0.09721647202968597,
|
| 1317 |
+
"learning_rate": 2.93638131693332e-05,
|
| 1318 |
+
"loss": 0.045375749468803406,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.1742936494695624,
|
| 1323 |
+
"grad_norm": 0.05705849081277847,
|
| 1324 |
+
"learning_rate": 2.9178726251918454e-05,
|
| 1325 |
+
"loss": 0.030177879333496093,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.1805745017026998,
|
| 1330 |
+
"grad_norm": 0.520708441734314,
|
| 1331 |
+
"learning_rate": 2.8993403405524617e-05,
|
| 1332 |
+
"loss": 0.038815781474113464,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.186855353935837,
|
| 1337 |
+
"grad_norm": 0.10496491193771362,
|
| 1338 |
+
"learning_rate": 2.8807855093395126e-05,
|
| 1339 |
+
"loss": 0.03444432914257049,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.1931362061689745,
|
| 1344 |
+
"grad_norm": 0.3142032027244568,
|
| 1345 |
+
"learning_rate": 2.8622091791503087e-05,
|
| 1346 |
+
"loss": 0.032994714379310605,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.199417058402112,
|
| 1351 |
+
"grad_norm": 0.0916871502995491,
|
| 1352 |
+
"learning_rate": 2.8436123987959822e-05,
|
| 1353 |
+
"loss": 0.0447750449180603,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.2056979106352492,
|
| 1358 |
+
"grad_norm": 1.1302578449249268,
|
| 1359 |
+
"learning_rate": 2.8249962182422757e-05,
|
| 1360 |
+
"loss": 0.05328472852706909,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.2119787628683867,
|
| 1365 |
+
"grad_norm": 0.5081734657287598,
|
| 1366 |
+
"learning_rate": 2.8063616885502535e-05,
|
| 1367 |
+
"loss": 0.06290997266769409,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.2182596151015241,
|
| 1372 |
+
"grad_norm": 0.13767530024051666,
|
| 1373 |
+
"learning_rate": 2.7877098618169682e-05,
|
| 1374 |
+
"loss": 0.04341446459293365,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.2245404673346614,
|
| 1379 |
+
"grad_norm": 5.364864349365234,
|
| 1380 |
+
"learning_rate": 2.769041791116054e-05,
|
| 1381 |
+
"loss": 0.0898139476776123,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.2308213195677988,
|
| 1386 |
+
"grad_norm": 0.21313656866550446,
|
| 1387 |
+
"learning_rate": 2.75035853043827e-05,
|
| 1388 |
+
"loss": 0.04104435741901398,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.2371021718009363,
|
| 1393 |
+
"grad_norm": 0.0669020265340805,
|
| 1394 |
+
"learning_rate": 2.7316611346319977e-05,
|
| 1395 |
+
"loss": 0.051208823919296265,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.2433830240340735,
|
| 1400 |
+
"grad_norm": 0.8109332919120789,
|
| 1401 |
+
"learning_rate": 2.7129506593436793e-05,
|
| 1402 |
+
"loss": 0.04516505599021912,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.249663876267211,
|
| 1407 |
+
"grad_norm": 0.10501603782176971,
|
| 1408 |
+
"learning_rate": 2.6942281609582205e-05,
|
| 1409 |
+
"loss": 0.03969334065914154,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.2559447285003484,
|
| 1414 |
+
"grad_norm": 0.22247670590877533,
|
| 1415 |
+
"learning_rate": 2.6754946965393434e-05,
|
| 1416 |
+
"loss": 0.059719407558441163,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.2622255807334857,
|
| 1421 |
+
"grad_norm": 0.10183948278427124,
|
| 1422 |
+
"learning_rate": 2.6567513237699087e-05,
|
| 1423 |
+
"loss": 0.06911430954933166,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.2685064329666231,
|
| 1428 |
+
"grad_norm": 0.15422746539115906,
|
| 1429 |
+
"learning_rate": 2.637999100892197e-05,
|
| 1430 |
+
"loss": 0.035851669311523435,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.2747872851997606,
|
| 1435 |
+
"grad_norm": 0.21719178557395935,
|
| 1436 |
+
"learning_rate": 2.6192390866481643e-05,
|
| 1437 |
+
"loss": 0.0500733494758606,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.2810681374328978,
|
| 1442 |
+
"grad_norm": 0.09273883700370789,
|
| 1443 |
+
"learning_rate": 2.6004723402196612e-05,
|
| 1444 |
+
"loss": 0.055965477228164674,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.2873489896660353,
|
| 1449 |
+
"grad_norm": 0.9721933603286743,
|
| 1450 |
+
"learning_rate": 2.5816999211686345e-05,
|
| 1451 |
+
"loss": 0.05600081086158752,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.2936298418991727,
|
| 1456 |
+
"grad_norm": 0.17564348876476288,
|
| 1457 |
+
"learning_rate": 2.562922889377307e-05,
|
| 1458 |
+
"loss": 0.03358841836452484,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.29991069413231,
|
| 1463 |
+
"grad_norm": 0.16629266738891602,
|
| 1464 |
+
"learning_rate": 2.544142304988331e-05,
|
| 1465 |
+
"loss": 0.028394874930381776,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.3061915463654474,
|
| 1470 |
+
"grad_norm": 0.6944901943206787,
|
| 1471 |
+
"learning_rate": 2.525359228344941e-05,
|
| 1472 |
+
"loss": 0.09227818250656128,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.3124723985985849,
|
| 1477 |
+
"grad_norm": 1.1704875230789185,
|
| 1478 |
+
"learning_rate": 2.50657471993108e-05,
|
| 1479 |
+
"loss": 0.06308795809745789,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.3187532508317221,
|
| 1484 |
+
"grad_norm": 0.103133924305439,
|
| 1485 |
+
"learning_rate": 2.487789840311529e-05,
|
| 1486 |
+
"loss": 0.022381560504436494,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.3250341030648596,
|
| 1491 |
+
"grad_norm": 0.4567931890487671,
|
| 1492 |
+
"learning_rate": 2.4690056500720276e-05,
|
| 1493 |
+
"loss": 0.06893195509910584,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.331314955297997,
|
| 1498 |
+
"grad_norm": 0.2637085020542145,
|
| 1499 |
+
"learning_rate": 2.450223209759392e-05,
|
| 1500 |
+
"loss": 0.03433942794799805,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.3375958075311343,
|
| 1505 |
+
"grad_norm": 0.09037027508020401,
|
| 1506 |
+
"learning_rate": 2.43144357982164e-05,
|
| 1507 |
+
"loss": 0.06628693342208862,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.3438766597642717,
|
| 1512 |
+
"grad_norm": 0.5666378736495972,
|
| 1513 |
+
"learning_rate": 2.4126678205481167e-05,
|
| 1514 |
+
"loss": 0.032296696305274965,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.3501575119974092,
|
| 1519 |
+
"grad_norm": 0.5473921298980713,
|
| 1520 |
+
"learning_rate": 2.39389699200963e-05,
|
| 1521 |
+
"loss": 0.04400340020656586,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.3564383642305464,
|
| 1526 |
+
"grad_norm": 0.44355085492134094,
|
| 1527 |
+
"learning_rate": 2.3751321539986e-05,
|
| 1528 |
+
"loss": 0.026118436455726625,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.3627192164636839,
|
| 1533 |
+
"grad_norm": 0.12921008467674255,
|
| 1534 |
+
"learning_rate": 2.3563743659692285e-05,
|
| 1535 |
+
"loss": 0.06551381945610046,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.3690000686968213,
|
| 1540 |
+
"grad_norm": 0.11393677443265915,
|
| 1541 |
+
"learning_rate": 2.3376246869776748e-05,
|
| 1542 |
+
"loss": 0.033758708834648134,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.3752809209299586,
|
| 1547 |
+
"grad_norm": 0.1252976655960083,
|
| 1548 |
+
"learning_rate": 2.3188841756222666e-05,
|
| 1549 |
+
"loss": 0.03352157473564148,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.381561773163096,
|
| 1554 |
+
"grad_norm": 0.3442012667655945,
|
| 1555 |
+
"learning_rate": 2.3001538899837304e-05,
|
| 1556 |
+
"loss": 0.02682451605796814,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.3878426253962335,
|
| 1561 |
+
"grad_norm": 0.05070227384567261,
|
| 1562 |
+
"learning_rate": 2.2814348875654547e-05,
|
| 1563 |
+
"loss": 0.048180288076400755,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.3941234776293707,
|
| 1568 |
+
"grad_norm": 0.04605748876929283,
|
| 1569 |
+
"learning_rate": 2.2627282252337827e-05,
|
| 1570 |
+
"loss": 0.04540911614894867,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.4004043298625082,
|
| 1575 |
+
"grad_norm": 0.40053147077560425,
|
| 1576 |
+
"learning_rate": 2.2440349591583403e-05,
|
| 1577 |
+
"loss": 0.03482065200805664,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.4066851820956456,
|
| 1582 |
+
"grad_norm": 0.10299244523048401,
|
| 1583 |
+
"learning_rate": 2.225356144752408e-05,
|
| 1584 |
+
"loss": 0.037977021932601926,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.4129660343287829,
|
| 1589 |
+
"grad_norm": 0.5996483564376831,
|
| 1590 |
+
"learning_rate": 2.2066928366133312e-05,
|
| 1591 |
+
"loss": 0.042719298601150514,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.4192468865619203,
|
| 1596 |
+
"grad_norm": 0.6384432911872864,
|
| 1597 |
+
"learning_rate": 2.188046088462979e-05,
|
| 1598 |
+
"loss": 0.03994882106781006,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.4255277387950578,
|
| 1603 |
+
"grad_norm": 0.05084218084812164,
|
| 1604 |
+
"learning_rate": 2.1694169530882496e-05,
|
| 1605 |
+
"loss": 0.04938451051712036,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.431808591028195,
|
| 1610 |
+
"grad_norm": 0.08827987313270569,
|
| 1611 |
+
"learning_rate": 2.1508064822816353e-05,
|
| 1612 |
+
"loss": 0.05632635354995728,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.4380894432613325,
|
| 1617 |
+
"grad_norm": 0.0749838724732399,
|
| 1618 |
+
"learning_rate": 2.1322157267818323e-05,
|
| 1619 |
+
"loss": 0.027722489833831788,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.44437029549447,
|
| 1624 |
+
"grad_norm": 0.08020087331533432,
|
| 1625 |
+
"learning_rate": 2.1136457362144192e-05,
|
| 1626 |
+
"loss": 0.030249813199043275,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.4506511477276072,
|
| 1631 |
+
"grad_norm": 0.15707388520240784,
|
| 1632 |
+
"learning_rate": 2.0950975590325994e-05,
|
| 1633 |
+
"loss": 0.02839776873588562,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.4569319999607446,
|
| 1638 |
+
"grad_norm": 0.22709378600120544,
|
| 1639 |
+
"learning_rate": 2.076572242457999e-05,
|
| 1640 |
+
"loss": 0.03741714060306549,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.463212852193882,
|
| 1645 |
+
"grad_norm": 0.3488466441631317,
|
| 1646 |
+
"learning_rate": 2.058070832421547e-05,
|
| 1647 |
+
"loss": 0.03505607545375824,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.4694937044270193,
|
| 1652 |
+
"grad_norm": 0.11174270510673523,
|
| 1653 |
+
"learning_rate": 2.0395943735044177e-05,
|
| 1654 |
+
"loss": 0.059834635257720946,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.4757745566601568,
|
| 1659 |
+
"grad_norm": 0.13519228994846344,
|
| 1660 |
+
"learning_rate": 2.0211439088790592e-05,
|
| 1661 |
+
"loss": 0.11914054155349732,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.4820554088932942,
|
| 1666 |
+
"grad_norm": 0.45727720856666565,
|
| 1667 |
+
"learning_rate": 2.0027204802502926e-05,
|
| 1668 |
+
"loss": 0.036192506551742554,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.4883362611264315,
|
| 1673 |
+
"grad_norm": 0.07157168537378311,
|
| 1674 |
+
"learning_rate": 1.9843251277964974e-05,
|
| 1675 |
+
"loss": 0.026058319211006164,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.494617113359569,
|
| 1680 |
+
"grad_norm": 0.08974478393793106,
|
| 1681 |
+
"learning_rate": 1.9659588901108887e-05,
|
| 1682 |
+
"loss": 0.04477571249008179,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.5008979655927064,
|
| 1687 |
+
"grad_norm": 0.05548376590013504,
|
| 1688 |
+
"learning_rate": 1.9476228041428722e-05,
|
| 1689 |
+
"loss": 0.03547537326812744,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.5071788178258436,
|
| 1694 |
+
"grad_norm": 0.20127072930335999,
|
| 1695 |
+
"learning_rate": 1.929317905139501e-05,
|
| 1696 |
+
"loss": 0.03266555070877075,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.513459670058981,
|
| 1701 |
+
"grad_norm": 0.3442573845386505,
|
| 1702 |
+
"learning_rate": 1.9110452265870294e-05,
|
| 1703 |
+
"loss": 0.03587172031402588,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.5197405222921185,
|
| 1708 |
+
"grad_norm": 0.22611093521118164,
|
| 1709 |
+
"learning_rate": 1.892805800152557e-05,
|
| 1710 |
+
"loss": 0.04896047115325928,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.5260213745252558,
|
| 1715 |
+
"grad_norm": 0.11305402964353561,
|
| 1716 |
+
"learning_rate": 1.874600655625785e-05,
|
| 1717 |
+
"loss": 0.02946520745754242,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.5323022267583932,
|
| 1722 |
+
"grad_norm": 0.15782569348812103,
|
| 1723 |
+
"learning_rate": 1.856430820860872e-05,
|
| 1724 |
+
"loss": 0.04476933181285858,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.5385830789915307,
|
| 1729 |
+
"grad_norm": 0.09827234596014023,
|
| 1730 |
+
"learning_rate": 1.838297321718408e-05,
|
| 1731 |
+
"loss": 0.0315948486328125,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.544863931224668,
|
| 1736 |
+
"grad_norm": 0.05839787423610687,
|
| 1737 |
+
"learning_rate": 1.8202011820074856e-05,
|
| 1738 |
+
"loss": 0.03258759379386902,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.5511447834578054,
|
| 1743 |
+
"grad_norm": 0.29925522208213806,
|
| 1744 |
+
"learning_rate": 1.8021434234279016e-05,
|
| 1745 |
+
"loss": 0.0613477885723114,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.5574256356909428,
|
| 1750 |
+
"grad_norm": 0.06653934717178345,
|
| 1751 |
+
"learning_rate": 1.7841250655124756e-05,
|
| 1752 |
+
"loss": 0.018268275260925292,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.56370648792408,
|
| 1757 |
+
"grad_norm": 0.49154847860336304,
|
| 1758 |
+
"learning_rate": 1.7661471255694792e-05,
|
| 1759 |
+
"loss": 0.0421527087688446,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.5699873401572177,
|
| 1764 |
+
"grad_norm": 0.08876429498195648,
|
| 1765 |
+
"learning_rate": 1.7482106186252055e-05,
|
| 1766 |
+
"loss": 0.04372061789035797,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.576268192390355,
|
| 1771 |
+
"grad_norm": 0.5607004165649414,
|
| 1772 |
+
"learning_rate": 1.7303165573666584e-05,
|
| 1773 |
+
"loss": 0.0424809455871582,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.5825490446234922,
|
| 1778 |
+
"grad_norm": 0.04403248429298401,
|
| 1779 |
+
"learning_rate": 1.7124659520843805e-05,
|
| 1780 |
+
"loss": 0.032825028896331786,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.58882989685663,
|
| 1785 |
+
"grad_norm": 0.17549766600131989,
|
| 1786 |
+
"learning_rate": 1.694659810615407e-05,
|
| 1787 |
+
"loss": 0.03757922351360321,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.5951107490897671,
|
| 1792 |
+
"grad_norm": 2.640838146209717,
|
| 1793 |
+
"learning_rate": 1.6768991382863663e-05,
|
| 1794 |
+
"loss": 0.042422395944595334,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.6013916013229044,
|
| 1799 |
+
"grad_norm": 0.323465496301651,
|
| 1800 |
+
"learning_rate": 1.6591849378567224e-05,
|
| 1801 |
+
"loss": 0.026809802651405333,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.607672453556042,
|
| 1806 |
+
"grad_norm": 0.7072569131851196,
|
| 1807 |
+
"learning_rate": 1.6415182094621548e-05,
|
| 1808 |
+
"loss": 0.0401191771030426,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.6139533057891793,
|
| 1813 |
+
"grad_norm": 0.10183272510766983,
|
| 1814 |
+
"learning_rate": 1.6238999505580944e-05,
|
| 1815 |
+
"loss": 0.033602625131607056,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.6202341580223165,
|
| 1820 |
+
"grad_norm": 0.7824767827987671,
|
| 1821 |
+
"learning_rate": 1.6063311558634054e-05,
|
| 1822 |
+
"loss": 0.038211938738822934,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.6265150102554542,
|
| 1827 |
+
"grad_norm": 0.5580938458442688,
|
| 1828 |
+
"learning_rate": 1.588812817304229e-05,
|
| 1829 |
+
"loss": 0.05440272688865662,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.6327958624885914,
|
| 1834 |
+
"grad_norm": 0.05703716725111008,
|
| 1835 |
+
"learning_rate": 1.5713459239579724e-05,
|
| 1836 |
+
"loss": 0.03235044479370117,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.6390767147217287,
|
| 1841 |
+
"grad_norm": 0.6257656216621399,
|
| 1842 |
+
"learning_rate": 1.5539314619974694e-05,
|
| 1843 |
+
"loss": 0.0461610347032547,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.6453575669548663,
|
| 1848 |
+
"grad_norm": 0.10126198083162308,
|
| 1849 |
+
"learning_rate": 1.5365704146353055e-05,
|
| 1850 |
+
"loss": 0.03510347306728363,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.6516384191880036,
|
| 1855 |
+
"grad_norm": 0.30662259459495544,
|
| 1856 |
+
"learning_rate": 1.5192637620682981e-05,
|
| 1857 |
+
"loss": 0.02332144379615784,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.6579192714211408,
|
| 1862 |
+
"grad_norm": 0.4968303143978119,
|
| 1863 |
+
"learning_rate": 1.5020124814221614e-05,
|
| 1864 |
+
"loss": 0.053898459672927855,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.6642001236542785,
|
| 1869 |
+
"grad_norm": 0.08359666168689728,
|
| 1870 |
+
"learning_rate": 1.484817546696335e-05,
|
| 1871 |
+
"loss": 0.03984528183937073,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.6704809758874157,
|
| 1876 |
+
"grad_norm": 0.10340140014886856,
|
| 1877 |
+
"learning_rate": 1.4676799287089954e-05,
|
| 1878 |
+
"loss": 0.042782101035118106,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.676761828120553,
|
| 1883 |
+
"grad_norm": 0.427079439163208,
|
| 1884 |
+
"learning_rate": 1.4506005950422402e-05,
|
| 1885 |
+
"loss": 0.06919811964035034,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.6830426803536906,
|
| 1890 |
+
"grad_norm": 0.26712024211883545,
|
| 1891 |
+
"learning_rate": 1.4335805099874611e-05,
|
| 1892 |
+
"loss": 0.02424788326025009,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.6893235325868279,
|
| 1897 |
+
"grad_norm": 0.040304891765117645,
|
| 1898 |
+
"learning_rate": 1.4166206344909027e-05,
|
| 1899 |
+
"loss": 0.03984163701534271,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.6956043848199651,
|
| 1904 |
+
"grad_norm": 0.2856264114379883,
|
| 1905 |
+
"learning_rate": 1.3997219260994022e-05,
|
| 1906 |
+
"loss": 0.030026772618293764,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.7018852370531028,
|
| 1911 |
+
"grad_norm": 0.0509999580681324,
|
| 1912 |
+
"learning_rate": 1.38288533890633e-05,
|
| 1913 |
+
"loss": 0.030039581656455993,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.70816608928624,
|
| 1918 |
+
"grad_norm": 1.1028673648834229,
|
| 1919 |
+
"learning_rate": 1.3661118234977261e-05,
|
| 1920 |
+
"loss": 0.028262528777122497,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.7144469415193773,
|
| 1925 |
+
"grad_norm": 0.7633464336395264,
|
| 1926 |
+
"learning_rate": 1.3494023268986211e-05,
|
| 1927 |
+
"loss": 0.027669858932495118,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.720727793752515,
|
| 1932 |
+
"grad_norm": 0.07768174260854721,
|
| 1933 |
+
"learning_rate": 1.3327577925195755e-05,
|
| 1934 |
+
"loss": 0.029605957865715026,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.7270086459856522,
|
| 1939 |
+
"grad_norm": 0.2423878312110901,
|
| 1940 |
+
"learning_rate": 1.3161791601034112e-05,
|
| 1941 |
+
"loss": 0.02931373715400696,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.7332894982187894,
|
| 1946 |
+
"grad_norm": 0.08954047411680222,
|
| 1947 |
+
"learning_rate": 1.2996673656721583e-05,
|
| 1948 |
+
"loss": 0.03215587437152863,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.739570350451927,
|
| 1953 |
+
"grad_norm": 0.24734511971473694,
|
| 1954 |
+
"learning_rate": 1.2832233414742013e-05,
|
| 1955 |
+
"loss": 0.0303827702999115,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.7458512026850643,
|
| 1960 |
+
"grad_norm": 0.14202502369880676,
|
| 1961 |
+
"learning_rate": 1.2668480159316478e-05,
|
| 1962 |
+
"loss": 0.022409896552562713,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.7521320549182016,
|
| 1967 |
+
"grad_norm": 0.06988602131605148,
|
| 1968 |
+
"learning_rate": 1.2505423135879155e-05,
|
| 1969 |
+
"loss": 0.034993928670883176,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.7584129071513392,
|
| 1974 |
+
"grad_norm": 0.3052821457386017,
|
| 1975 |
+
"learning_rate": 1.2343071550555232e-05,
|
| 1976 |
+
"loss": 0.022670230269432066,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.7646937593844765,
|
| 1981 |
+
"grad_norm": 0.07533857971429825,
|
| 1982 |
+
"learning_rate": 1.21814345696412e-05,
|
| 1983 |
+
"loss": 0.027903926372528077,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.7709746116176137,
|
| 1988 |
+
"grad_norm": 0.04599501192569733,
|
| 1989 |
+
"learning_rate": 1.2020521319087313e-05,
|
| 1990 |
+
"loss": 0.022192281484603883,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.7772554638507514,
|
| 1995 |
+
"grad_norm": 0.18866924941539764,
|
| 1996 |
+
"learning_rate": 1.1860340883982335e-05,
|
| 1997 |
+
"loss": 0.026816827058792115,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.7835363160838886,
|
| 2002 |
+
"grad_norm": 0.07341870665550232,
|
| 2003 |
+
"learning_rate": 1.1700902308040605e-05,
|
| 2004 |
+
"loss": 0.021274025738239288,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.7898171683170259,
|
| 2009 |
+
"grad_norm": 0.3347774147987366,
|
| 2010 |
+
"learning_rate": 1.1542214593091419e-05,
|
| 2011 |
+
"loss": 0.03206228017807007,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.7960980205501635,
|
| 2016 |
+
"grad_norm": 0.06323454529047012,
|
| 2017 |
+
"learning_rate": 1.1384286698570837e-05,
|
| 2018 |
+
"loss": 0.02469037175178528,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.8023788727833008,
|
| 2023 |
+
"grad_norm": 0.11099821329116821,
|
| 2024 |
+
"learning_rate": 1.1227127541015775e-05,
|
| 2025 |
+
"loss": 0.02508939504623413,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.808659725016438,
|
| 2030 |
+
"grad_norm": 0.10954418778419495,
|
| 2031 |
+
"learning_rate": 1.1070745993560618e-05,
|
| 2032 |
+
"loss": 0.053559088706970216,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.8149405772495757,
|
| 2037 |
+
"grad_norm": 0.48120102286338806,
|
| 2038 |
+
"learning_rate": 1.0915150885436235e-05,
|
| 2039 |
+
"loss": 0.03702453076839447,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.821221429482713,
|
| 2044 |
+
"grad_norm": 0.5513730645179749,
|
| 2045 |
+
"learning_rate": 1.0760351001471508e-05,
|
| 2046 |
+
"loss": 0.05739341974258423,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.8275022817158502,
|
| 2051 |
+
"grad_norm": 0.05452940985560417,
|
| 2052 |
+
"learning_rate": 1.0606355081597302e-05,
|
| 2053 |
+
"loss": 0.032155710458755496,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.8337831339489878,
|
| 2058 |
+
"grad_norm": 0.17276941239833832,
|
| 2059 |
+
"learning_rate": 1.0453171820353037e-05,
|
| 2060 |
+
"loss": 0.028540417551994324,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.840063986182125,
|
| 2065 |
+
"grad_norm": 0.35029998421669006,
|
| 2066 |
+
"learning_rate": 1.0300809866395825e-05,
|
| 2067 |
+
"loss": 0.026354405283927917,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 1.8463448384152623,
|
| 2072 |
+
"grad_norm": 0.17696386575698853,
|
| 2073 |
+
"learning_rate": 1.0149277822012116e-05,
|
| 2074 |
+
"loss": 0.04181792438030243,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 1.8526256906484,
|
| 2079 |
+
"grad_norm": 0.07578734308481216,
|
| 2080 |
+
"learning_rate": 9.99858424263204e-06,
|
| 2081 |
+
"loss": 0.024270452558994293,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.8589065428815372,
|
| 2086 |
+
"grad_norm": 0.12244296818971634,
|
| 2087 |
+
"learning_rate": 9.848737636346403e-06,
|
| 2088 |
+
"loss": 0.03891803920269012,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 1.8651873951146745,
|
| 2093 |
+
"grad_norm": 0.14096958935260773,
|
| 2094 |
+
"learning_rate": 9.699746463426276e-06,
|
| 2095 |
+
"loss": 0.05135524272918701,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.8714682473478121,
|
| 2100 |
+
"grad_norm": 0.09481539577245712,
|
| 2101 |
+
"learning_rate": 9.551619135845352e-06,
|
| 2102 |
+
"loss": 0.032832151651382445,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.8777490995809494,
|
| 2107 |
+
"grad_norm": 0.07355910539627075,
|
| 2108 |
+
"learning_rate": 9.404364016805009e-06,
|
| 2109 |
+
"loss": 0.016955824196338655,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.8840299518140866,
|
| 2114 |
+
"grad_norm": 0.5096582174301147,
|
| 2115 |
+
"learning_rate": 9.257989420262151e-06,
|
| 2116 |
+
"loss": 0.024562205374240875,
|
| 2117 |
+
"step": 6000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.8903108040472243,
|
| 2121 |
+
"grad_norm": 1.0709404945373535,
|
| 2122 |
+
"learning_rate": 9.112503610459751e-06,
|
| 2123 |
+
"loss": 0.039287084341049196,
|
| 2124 |
+
"step": 6020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.8965916562803615,
|
| 2128 |
+
"grad_norm": 2.2170796394348145,
|
| 2129 |
+
"learning_rate": 8.96791480146031e-06,
|
| 2130 |
+
"loss": 0.03112347424030304,
|
| 2131 |
+
"step": 6040
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 1.9028725085134988,
|
| 2135 |
+
"grad_norm": 0.04912661388516426,
|
| 2136 |
+
"learning_rate": 8.824231156682063e-06,
|
| 2137 |
+
"loss": 0.05167080163955688,
|
| 2138 |
+
"step": 6060
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 1.9091533607466364,
|
| 2142 |
+
"grad_norm": 0.06297567486763,
|
| 2143 |
+
"learning_rate": 8.681460788438087e-06,
|
| 2144 |
+
"loss": 0.045577284693717954,
|
| 2145 |
+
"step": 6080
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.9154342129797737,
|
| 2149 |
+
"grad_norm": 0.11223562806844711,
|
| 2150 |
+
"learning_rate": 8.53961175747828e-06,
|
| 2151 |
+
"loss": 0.021874111890792847,
|
| 2152 |
+
"step": 6100
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 1.921715065212911,
|
| 2156 |
+
"grad_norm": 0.18214787542819977,
|
| 2157 |
+
"learning_rate": 8.398692072534258e-06,
|
| 2158 |
+
"loss": 0.029563134908676146,
|
| 2159 |
+
"step": 6120
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 1.9279959174460486,
|
| 2163 |
+
"grad_norm": 0.26750287413597107,
|
| 2164 |
+
"learning_rate": 8.258709689867203e-06,
|
| 2165 |
+
"loss": 0.02882084548473358,
|
| 2166 |
+
"step": 6140
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 1.9342767696791858,
|
| 2170 |
+
"grad_norm": 0.36355850100517273,
|
| 2171 |
+
"learning_rate": 8.119672512818608e-06,
|
| 2172 |
+
"loss": 0.025364571809768678,
|
| 2173 |
+
"step": 6160
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 1.9405576219123233,
|
| 2177 |
+
"grad_norm": 0.7907804846763611,
|
| 2178 |
+
"learning_rate": 7.981588391364098e-06,
|
| 2179 |
+
"loss": 0.031220081448554992,
|
| 2180 |
+
"step": 6180
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.9468384741454607,
|
| 2184 |
+
"grad_norm": 0.7984992861747742,
|
| 2185 |
+
"learning_rate": 7.844465121670233e-06,
|
| 2186 |
+
"loss": 0.04252537488937378,
|
| 2187 |
+
"step": 6200
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.953119326378598,
|
| 2191 |
+
"grad_norm": 0.041818082332611084,
|
| 2192 |
+
"learning_rate": 7.708310445654292e-06,
|
| 2193 |
+
"loss": 0.04662574529647827,
|
| 2194 |
+
"step": 6220
|
| 2195 |
+
}
|
| 2196 |
+
],
|
| 2197 |
+
"logging_steps": 20,
|
| 2198 |
+
"max_steps": 8362,
|
| 2199 |
+
"num_input_tokens_seen": 0,
|
| 2200 |
+
"num_train_epochs": 3,
|
| 2201 |
+
"save_steps": 1000000000,
|
| 2202 |
+
"stateful_callbacks": {
|
| 2203 |
+
"TrainerControl": {
|
| 2204 |
+
"args": {
|
| 2205 |
+
"should_epoch_stop": false,
|
| 2206 |
+
"should_evaluate": false,
|
| 2207 |
+
"should_log": false,
|
| 2208 |
+
"should_save": true,
|
| 2209 |
+
"should_training_stop": false
|
| 2210 |
+
},
|
| 2211 |
+
"attributes": {}
|
| 2212 |
+
}
|
| 2213 |
+
},
|
| 2214 |
+
"total_flos": 1.969628228368128e+16,
|
| 2215 |
+
"train_batch_size": 1,
|
| 2216 |
+
"trial_name": null,
|
| 2217 |
+
"trial_params": null
|
| 2218 |
+
}
|