init

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.38.1",
+  "use_cache": false
+}

model-00001-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecf09a137dc9370ff2fdaccfd224f3cc76e740dbbc27d0e55b0dbd98e9b635f0
+size 4933725040

model-00002-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17b803b1cdf74d084c58d00b9bafc7a144fbb6959b1e6b35923b450dccfa58e6
+size 4915916176

model-00003-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d6de738bcfbb7540234185024060d183ff7c9998b328337791a6e7e065d692a
+size 4999819336

model-00004-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21d58de47a649d1a263c2649a14ccad3841cecf942f0e00e94ee3581ec4b8fd0
+size 4915916184

model-00005-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddbbe94a61f5b5a21bfea5438f1390ff9b764fb4b5aeafba4545679a927481b4
+size 1913728736

model.safetensors.index.json

@@ -0,0 +1,442 @@
+{
+  "metadata": {
+    "total_size": 21679054848
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00005-of-00005.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00005.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00003-of-00005.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.36.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.37.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.38.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.39.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.40.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.40.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.41.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.42.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.mlp.down_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.mlp.up_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.43.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.44.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.44.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.44.mlp.gate_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.44.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.44.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.44.self_attn.k_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.44.self_attn.o_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.44.self_attn.q_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.44.self_attn.v_proj.weight": "model-00004-of-00005.safetensors",
+    "model.layers.45.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.45.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.46.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.mlp.down_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.mlp.gate_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.mlp.up_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.self_attn.k_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.self_attn.o_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.self_attn.q_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.47.self_attn.v_proj.weight": "model-00005-of-00005.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00005.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00005.safetensors",
+    "model.norm.weight": "model-00005-of-00005.safetensors"
+  }
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0ac66640a4707dbadff443d3e179e68b78039bc3349dece5d791b297ffd59d1
+size 1480592220

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:386fcc8cc1089aade9450d86fb239ea3483f455fd2d78d8378645feecfec9d69
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da9ab3b1cd0eedf96c680680bd28b90c2100079af79744dc3cfc9e2f8b824c6f
+size 1064

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe1f1df46f5bbbe47211af84f7ec05e86f38a4e5651f5dc99de7d63e2b636b64
+size 750737

tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": true,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 4096,
+  "pad_token": "</s>",
+  "padding_side": "left",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

trainer_state.json

@@ -0,0 +1,2891 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.401616943554929,
+  "eval_steps": 2,
+  "global_step": 59040,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.5,
+      "learning_rate": 3.264565858082068e-06,
+      "loss": 9.7328,
+      "step": 144
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.75,
+      "learning_rate": 6.529131716164136e-06,
+      "loss": 8.8374,
+      "step": 288
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.793697574246204e-06,
+      "loss": 7.1244,
+      "step": 432
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.3058263432328271e-05,
+      "loss": 6.2564,
+      "step": 576
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.6322829290410337e-05,
+      "loss": 5.7834,
+      "step": 720
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.958739514849241e-05,
+      "loss": 5.3786,
+      "step": 864
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.2851961006574474e-05,
+      "loss": 4.8869,
+      "step": 1008
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.6116526864656543e-05,
+      "loss": 4.4858,
+      "step": 1152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.9381092722738608e-05,
+      "loss": 4.1009,
+      "step": 1296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.264565858082067e-05,
+      "loss": 3.7652,
+      "step": 1440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.5910224438902745e-05,
+      "loss": 3.4421,
+      "step": 1584
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.917479029698482e-05,
+      "loss": 3.2573,
+      "step": 1728
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.2439356155066876e-05,
+      "loss": 3.0644,
+      "step": 1872
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.570392201314895e-05,
+      "loss": 2.9334,
+      "step": 2016
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.125,
+      "learning_rate": 4.896848787123102e-05,
+      "loss": 2.8316,
+      "step": 2160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.2233053729313085e-05,
+      "loss": 2.7194,
+      "step": 2304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.549761958739516e-05,
+      "loss": 2.6336,
+      "step": 2448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.8762185445477216e-05,
+      "loss": 2.5866,
+      "step": 2592
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.202675130355929e-05,
+      "loss": 2.5192,
+      "step": 2736
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.15625,
+      "learning_rate": 6.529131716164135e-05,
+      "loss": 2.4692,
+      "step": 2880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.855588301972342e-05,
+      "loss": 2.4297,
+      "step": 3024
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.182044887780549e-05,
+      "loss": 2.4255,
+      "step": 3168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.508501473588756e-05,
+      "loss": 2.3646,
+      "step": 3312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.834958059396963e-05,
+      "loss": 2.3059,
+      "step": 3456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.161414645205169e-05,
+      "loss": 2.2926,
+      "step": 3600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.487871231013375e-05,
+      "loss": 2.2704,
+      "step": 3744
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.814327816821582e-05,
+      "loss": 2.251,
+      "step": 3888
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.14078440262979e-05,
+      "loss": 2.2393,
+      "step": 4032
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.467240988437997e-05,
+      "loss": 2.2282,
+      "step": 4176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.625,
+      "learning_rate": 9.793697574246204e-05,
+      "loss": 2.2133,
+      "step": 4320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.999996591298748e-05,
+      "loss": 2.1821,
+      "step": 4464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.99995290562682e-05,
+      "loss": 2.1851,
+      "step": 4608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.999858894492333e-05,
+      "loss": 2.1736,
+      "step": 4752
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.99971455884153e-05,
+      "loss": 2.1474,
+      "step": 4896
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.999519900127174e-05,
+      "loss": 2.144,
+      "step": 5040
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5,
+      "learning_rate": 9.999274920308544e-05,
+      "loss": 2.1284,
+      "step": 5184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.998979621851405e-05,
+      "loss": 2.1219,
+      "step": 5328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.998634007727992e-05,
+      "loss": 2.1004,
+      "step": 5472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.625,
+      "learning_rate": 9.998238081416977e-05,
+      "loss": 2.1148,
+      "step": 5616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.99779184690343e-05,
+      "loss": 2.119,
+      "step": 5760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.997295308678789e-05,
+      "loss": 2.1017,
+      "step": 5904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.996748471740802e-05,
+      "loss": 2.098,
+      "step": 6048
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.996151341593482e-05,
+      "loss": 2.0888,
+      "step": 6192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.995503924247063e-05,
+      "loss": 2.0885,
+      "step": 6336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.994806226217915e-05,
+      "loss": 2.0654,
+      "step": 6480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.994058254528505e-05,
+      "loss": 2.0888,
+      "step": 6624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.993260016707303e-05,
+      "loss": 2.0717,
+      "step": 6768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.992411520788724e-05,
+      "loss": 2.0513,
+      "step": 6912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.99151277531304e-05,
+      "loss": 2.0695,
+      "step": 7056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.990563789326291e-05,
+      "loss": 2.0612,
+      "step": 7200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.989564572380199e-05,
+      "loss": 2.0577,
+      "step": 7344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.98851513453207e-05,
+      "loss": 2.0453,
+      "step": 7488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.987415486344694e-05,
+      "loss": 2.0381,
+      "step": 7632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.986265638886232e-05,
+      "loss": 2.0528,
+      "step": 7776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.985065603730119e-05,
+      "loss": 2.0414,
+      "step": 7920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.98381539295493e-05,
+      "loss": 2.0388,
+      "step": 8064
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.98251501914427e-05,
+      "loss": 2.0096,
+      "step": 8208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.981164495386647e-05,
+      "loss": 2.023,
+      "step": 8352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.979763835275338e-05,
+      "loss": 2.0058,
+      "step": 8496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.375,
+      "learning_rate": 9.978313052908243e-05,
+      "loss": 2.0333,
+      "step": 8640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.976812162887765e-05,
+      "loss": 2.0387,
+      "step": 8784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.975261180320638e-05,
+      "loss": 2.0307,
+      "step": 8928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.973660120817796e-05,
+      "loss": 2.0153,
+      "step": 9072
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.9720090004942e-05,
+      "loss": 2.0324,
+      "step": 9216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.970307835968687e-05,
+      "loss": 2.0118,
+      "step": 9360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.968556644363793e-05,
+      "loss": 2.0003,
+      "step": 9504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.966755443305595e-05,
+      "loss": 2.0019,
+      "step": 9648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.964904250923517e-05,
+      "loss": 2.0126,
+      "step": 9792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.963003085850155e-05,
+      "loss": 1.9971,
+      "step": 9936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.961051967221096e-05,
+      "loss": 1.9832,
+      "step": 10080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.959050914674708e-05,
+      "loss": 1.9898,
+      "step": 10224
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.956999948351967e-05,
+      "loss": 1.9954,
+      "step": 10368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.954899088896226e-05,
+      "loss": 1.9928,
+      "step": 10512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.952748357453034e-05,
+      "loss": 1.9925,
+      "step": 10656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.950547775669903e-05,
+      "loss": 1.9898,
+      "step": 10800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.948297365696104e-05,
+      "loss": 2.0034,
+      "step": 10944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.945997150182432e-05,
+      "loss": 2.0083,
+      "step": 11088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.943647152280988e-05,
+      "loss": 1.9859,
+      "step": 11232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.941247395644942e-05,
+      "loss": 1.9656,
+      "step": 11376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.938797904428294e-05,
+      "loss": 1.9793,
+      "step": 11520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.936298703285629e-05,
+      "loss": 1.9857,
+      "step": 11664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.933749817371878e-05,
+      "loss": 1.9909,
+      "step": 11808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.931151272342054e-05,
+      "loss": 1.9626,
+      "step": 11952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.928503094351e-05,
+      "loss": 1.9937,
+      "step": 12096
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.92580531005312e-05,
+      "loss": 1.9876,
+      "step": 12240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.923057946602125e-05,
+      "loss": 1.9967,
+      "step": 12384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.920261031650738e-05,
+      "loss": 1.9712,
+      "step": 12528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0,
+      "learning_rate": 9.917414593350432e-05,
+      "loss": 1.9782,
+      "step": 12672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.914518660351142e-05,
+      "loss": 1.9822,
+      "step": 12816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.911573261800977e-05,
+      "loss": 1.9826,
+      "step": 12960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.908578427345925e-05,
+      "loss": 1.9701,
+      "step": 13104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.625,
+      "learning_rate": 9.905534187129555e-05,
+      "loss": 1.9822,
+      "step": 13248
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.902440571792714e-05,
+      "loss": 1.9728,
+      "step": 13392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.899297612473225e-05,
+      "loss": 1.968,
+      "step": 13536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.896105340805557e-05,
+      "loss": 1.9723,
+      "step": 13680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.892863788920523e-05,
+      "loss": 1.9676,
+      "step": 13824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.125,
+      "learning_rate": 9.889572989444953e-05,
+      "loss": 1.9499,
+      "step": 13968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.886232975501355e-05,
+      "loss": 1.985,
+      "step": 14112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.8828437807076e-05,
+      "loss": 1.9802,
+      "step": 14256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.879405439176566e-05,
+      "loss": 1.9356,
+      "step": 14400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.875917985515805e-05,
+      "loss": 1.9518,
+      "step": 14544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.872381454827191e-05,
+      "loss": 1.9851,
+      "step": 14688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.868795882706569e-05,
+      "loss": 1.9658,
+      "step": 14832
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.86516130524339e-05,
+      "loss": 1.9648,
+      "step": 14976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.861477759020363e-05,
+      "loss": 1.9852,
+      "step": 15120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.857745281113069e-05,
+      "loss": 1.9768,
+      "step": 15264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.853963909089594e-05,
+      "loss": 1.958,
+      "step": 15408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.850133681010159e-05,
+      "loss": 1.9697,
+      "step": 15552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.846254635426724e-05,
+      "loss": 1.976,
+      "step": 15696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.842326811382612e-05,
+      "loss": 1.9644,
+      "step": 15840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.838350248412107e-05,
+      "loss": 1.9412,
+      "step": 15984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.83432498654006e-05,
+      "loss": 1.9672,
+      "step": 16128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.830251066281489e-05,
+      "loss": 1.9348,
+      "step": 16272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.826128528641157e-05,
+      "loss": 1.9759,
+      "step": 16416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.821957415113186e-05,
+      "loss": 1.9671,
+      "step": 16560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.817737767680608e-05,
+      "loss": 1.9524,
+      "step": 16704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.813469628814969e-05,
+      "loss": 1.9456,
+      "step": 16848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.809153041475883e-05,
+      "loss": 1.9576,
+      "step": 16992
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.804788049110615e-05,
+      "loss": 1.9802,
+      "step": 17136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.800374695653627e-05,
+      "loss": 1.9562,
+      "step": 17280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.795913025526148e-05,
+      "loss": 1.9342,
+      "step": 17424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.791403083635726e-05,
+      "loss": 1.9553,
+      "step": 17568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.786844915375768e-05,
+      "loss": 1.9813,
+      "step": 17712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.375,
+      "learning_rate": 9.782238566625094e-05,
+      "loss": 1.9396,
+      "step": 17856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.777584083747466e-05,
+      "loss": 1.9618,
+      "step": 18000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.772881513591124e-05,
+      "loss": 1.9451,
+      "step": 18144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.768130903488322e-05,
+      "loss": 1.9672,
+      "step": 18288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.763332301254835e-05,
+      "loss": 1.9764,
+      "step": 18432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.7584857551895e-05,
+      "loss": 1.9538,
+      "step": 18576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.753591314073705e-05,
+      "loss": 1.9473,
+      "step": 18720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.748649027170918e-05,
+      "loss": 1.9756,
+      "step": 18864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.743658944226185e-05,
+      "loss": 1.9427,
+      "step": 19008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.738621115465624e-05,
+      "loss": 1.9448,
+      "step": 19152
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.733535591595924e-05,
+      "loss": 1.9604,
+      "step": 19296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.728402423803836e-05,
+      "loss": 1.9488,
+      "step": 19440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.723221663755657e-05,
+      "loss": 1.9586,
+      "step": 19584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.717993363596705e-05,
+      "loss": 1.9509,
+      "step": 19728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.375,
+      "learning_rate": 9.712717575950802e-05,
+      "loss": 1.9435,
+      "step": 19872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.707394353919737e-05,
+      "loss": 1.9581,
+      "step": 20016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.70202375108274e-05,
+      "loss": 1.9395,
+      "step": 20160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.696605821495932e-05,
+      "loss": 1.9643,
+      "step": 20304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.691140619691789e-05,
+      "loss": 1.9318,
+      "step": 20448
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.40625,
+      "learning_rate": 9.685628200678592e-05,
+      "loss": 1.9463,
+      "step": 20592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.680068619939872e-05,
+      "loss": 1.9573,
+      "step": 20736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.674461933433849e-05,
+      "loss": 1.9454,
+      "step": 20880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.668808197592871e-05,
+      "loss": 1.9413,
+      "step": 21024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.663107469322852e-05,
+      "loss": 1.9387,
+      "step": 21168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3828125,
+      "learning_rate": 9.657359806002689e-05,
+      "loss": 1.9389,
+      "step": 21312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.651565265483687e-05,
+      "loss": 1.9477,
+      "step": 21456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.645723906088983e-05,
+      "loss": 1.9358,
+      "step": 21600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.25,
+      "learning_rate": 9.639835786612957e-05,
+      "loss": 1.9403,
+      "step": 21744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.125,
+      "learning_rate": 9.633900966320631e-05,
+      "loss": 1.9274,
+      "step": 21888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.627919504947084e-05,
+      "loss": 1.9236,
+      "step": 22032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.621891462696843e-05,
+      "loss": 1.947,
+      "step": 22176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.615816900243287e-05,
+      "loss": 1.9404,
+      "step": 22320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.60969587872802e-05,
+      "loss": 1.9458,
+      "step": 22464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.603528459760274e-05,
+      "loss": 1.9203,
+      "step": 22608
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.597314705416274e-05,
+      "loss": 1.943,
+      "step": 22752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.59105467823862e-05,
+      "loss": 1.9428,
+      "step": 22896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.584748441235663e-05,
+      "loss": 1.9281,
+      "step": 23040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.578396057880854e-05,
+      "loss": 1.9449,
+      "step": 23184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.571997592112126e-05,
+      "loss": 1.9509,
+      "step": 23328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.565553108331231e-05,
+      "loss": 1.9423,
+      "step": 23472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.559062671403109e-05,
+      "loss": 1.9272,
+      "step": 23616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.552526346655222e-05,
+      "loss": 1.9438,
+      "step": 23760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.545944199876904e-05,
+      "loss": 1.9523,
+      "step": 23904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.539316297318695e-05,
+      "loss": 1.9306,
+      "step": 24048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.532642705691674e-05,
+      "loss": 1.9391,
+      "step": 24192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.525923492166792e-05,
+      "loss": 1.9723,
+      "step": 24336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.519158724374193e-05,
+      "loss": 1.9303,
+      "step": 24480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.125,
+      "learning_rate": 9.51234847040253e-05,
+      "loss": 1.9424,
+      "step": 24624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.505492798798286e-05,
+      "loss": 1.9572,
+      "step": 24768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.49859177856508e-05,
+      "loss": 1.9425,
+      "step": 24912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.491645479162971e-05,
+      "loss": 1.9466,
+      "step": 25056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.484653970507766e-05,
+      "loss": 1.9457,
+      "step": 25200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.477617322970309e-05,
+      "loss": 1.9162,
+      "step": 25344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.470535607375775e-05,
+      "loss": 1.9428,
+      "step": 25488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.463408895002958e-05,
+      "loss": 1.9496,
+      "step": 25632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.456237257583555e-05,
+      "loss": 1.9268,
+      "step": 25776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.449020767301435e-05,
+      "loss": 1.934,
+      "step": 25920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2421875,
+      "learning_rate": 9.441759496791929e-05,
+      "loss": 1.919,
+      "step": 26064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.434453519141085e-05,
+      "loss": 1.9358,
+      "step": 26208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.427102907884933e-05,
+      "loss": 1.9278,
+      "step": 26352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.125,
+      "learning_rate": 9.419707737008754e-05,
+      "loss": 1.9417,
+      "step": 26496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.412268080946327e-05,
+      "loss": 1.9371,
+      "step": 26640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.404784014579186e-05,
+      "loss": 1.9495,
+      "step": 26784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.397255613235862e-05,
+      "loss": 1.9303,
+      "step": 26928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.38968295269112e-05,
+      "loss": 1.9501,
+      "step": 27072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.38206610916521e-05,
+      "loss": 1.9409,
+      "step": 27216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.374405159323087e-05,
+      "loss": 1.9442,
+      "step": 27360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.36670018027365e-05,
+      "loss": 1.9354,
+      "step": 27504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.358951249568952e-05,
+      "loss": 1.9349,
+      "step": 27648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.351158445203434e-05,
+      "loss": 1.9287,
+      "step": 27792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.343321845613132e-05,
+      "loss": 1.9313,
+      "step": 27936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.335441529674888e-05,
+      "loss": 1.9414,
+      "step": 28080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.327517576705558e-05,
+      "loss": 1.9269,
+      "step": 28224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.319550066461214e-05,
+      "loss": 1.9202,
+      "step": 28368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.311539079136336e-05,
+      "loss": 1.9235,
+      "step": 28512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.303484695363016e-05,
+      "loss": 1.9439,
+      "step": 28656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.295386996210133e-05,
+      "loss": 1.8996,
+      "step": 28800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.287246063182546e-05,
+      "loss": 1.9243,
+      "step": 28944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.279061978220272e-05,
+      "loss": 1.9174,
+      "step": 29088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.270834823697658e-05,
+      "loss": 1.9144,
+      "step": 29232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.262564682422554e-05,
+      "loss": 1.9212,
+      "step": 29376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.254251637635485e-05,
+      "loss": 1.9362,
+      "step": 29520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.245895773008799e-05,
+      "loss": 1.9275,
+      "step": 29664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.237497172645843e-05,
+      "loss": 1.9486,
+      "step": 29808
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.229055921080101e-05,
+      "loss": 1.9109,
+      "step": 29952
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.220572103274352e-05,
+      "loss": 1.9268,
+      "step": 30096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.212045804619809e-05,
+      "loss": 1.9211,
+      "step": 30240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.203477110935267e-05,
+      "loss": 1.9151,
+      "step": 30384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.194866108466233e-05,
+      "loss": 1.928,
+      "step": 30528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.18621288388406e-05,
+      "loss": 1.9282,
+      "step": 30672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.177517524285074e-05,
+      "loss": 1.9251,
+      "step": 30816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.168780117189696e-05,
+      "loss": 1.9425,
+      "step": 30960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.160000750541569e-05,
+      "loss": 1.9576,
+      "step": 31104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.125,
+      "learning_rate": 9.15117951270666e-05,
+      "loss": 1.9337,
+      "step": 31248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.14231649247238e-05,
+      "loss": 1.9047,
+      "step": 31392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.13341177904669e-05,
+      "loss": 1.9391,
+      "step": 31536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3671875,
+      "learning_rate": 9.124465462057196e-05,
+      "loss": 1.9066,
+      "step": 31680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.115477631550258e-05,
+      "loss": 1.9169,
+      "step": 31824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.106448377990068e-05,
+      "loss": 1.9238,
+      "step": 31968
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.097377792257759e-05,
+      "loss": 1.9373,
+      "step": 32112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.088265965650476e-05,
+      "loss": 1.9265,
+      "step": 32256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.079112989880455e-05,
+      "loss": 1.9111,
+      "step": 32400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.069918957074118e-05,
+      "loss": 1.9345,
+      "step": 32544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.060683959771124e-05,
+      "loss": 1.9157,
+      "step": 32688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.051408090923449e-05,
+      "loss": 1.9033,
+      "step": 32832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.042091443894456e-05,
+      "loss": 1.9159,
+      "step": 32976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.032734112457937e-05,
+      "loss": 1.9182,
+      "step": 33120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.02333619079719e-05,
+      "loss": 1.9231,
+      "step": 33264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.013897773504052e-05,
+      "loss": 1.9234,
+      "step": 33408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.004418955577965e-05,
+      "loss": 1.9147,
+      "step": 33552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.994899832425005e-05,
+      "loss": 1.924,
+      "step": 33696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.985340499856933e-05,
+      "loss": 1.9249,
+      "step": 33840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.975741054090216e-05,
+      "loss": 1.9018,
+      "step": 33984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.96610159174508e-05,
+      "loss": 1.9359,
+      "step": 34128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.956422209844514e-05,
+      "loss": 1.9366,
+      "step": 34272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.946703005813312e-05,
+      "loss": 1.921,
+      "step": 34416
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.936944077477084e-05,
+      "loss": 1.916,
+      "step": 34560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.92714552306127e-05,
+      "loss": 1.9121,
+      "step": 34704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.917307441190158e-05,
+      "loss": 1.907,
+      "step": 34848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.907429930885882e-05,
+      "loss": 1.9058,
+      "step": 34992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.897513091567435e-05,
+      "loss": 1.924,
+      "step": 35136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.88755702304966e-05,
+      "loss": 1.9397,
+      "step": 35280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.877561825542256e-05,
+      "loss": 1.9426,
+      "step": 35424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.867527599648755e-05,
+      "loss": 1.9151,
+      "step": 35568
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.85745444636552e-05,
+      "loss": 1.9062,
+      "step": 35712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.847342467080729e-05,
+      "loss": 1.9036,
+      "step": 35856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.837191763573343e-05,
+      "loss": 1.9183,
+      "step": 36000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.827002438012095e-05,
+      "loss": 1.9288,
+      "step": 36144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.816774592954458e-05,
+      "loss": 1.9296,
+      "step": 36288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.806508331345609e-05,
+      "loss": 1.9343,
+      "step": 36432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.79620375651739e-05,
+      "loss": 1.9268,
+      "step": 36576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.785860972187279e-05,
+      "loss": 1.9331,
+      "step": 36720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.775480082457336e-05,
+      "loss": 1.9028,
+      "step": 36864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.765061191813165e-05,
+      "loss": 1.9192,
+      "step": 37008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.375,
+      "learning_rate": 8.754604405122848e-05,
+      "loss": 1.9208,
+      "step": 37152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.744109827635902e-05,
+      "loss": 1.9039,
+      "step": 37296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.733577564982218e-05,
+      "loss": 1.9271,
+      "step": 37440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.723007723170988e-05,
+      "loss": 1.9236,
+      "step": 37584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.712400408589654e-05,
+      "loss": 1.9426,
+      "step": 37728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.701755728002823e-05,
+      "loss": 1.9107,
+      "step": 37872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.125,
+      "learning_rate": 8.691073788551196e-05,
+      "loss": 1.9313,
+      "step": 38016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.680354697750499e-05,
+      "loss": 1.9089,
+      "step": 38160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.669598563490385e-05,
+      "loss": 1.9294,
+      "step": 38304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.658805494033365e-05,
+      "loss": 1.9045,
+      "step": 38448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.647975598013698e-05,
+      "loss": 1.9113,
+      "step": 38592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.63710898443632e-05,
+      "loss": 1.9295,
+      "step": 38736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.626205762675732e-05,
+      "loss": 1.9194,
+      "step": 38880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.615266042474904e-05,
+      "loss": 1.9299,
+      "step": 39024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.604289933944165e-05,
+      "loss": 1.9278,
+      "step": 39168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.593277547560108e-05,
+      "loss": 1.9234,
+      "step": 39312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.582228994164463e-05,
+      "loss": 1.9134,
+      "step": 39456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.571144384962991e-05,
+      "loss": 1.9048,
+      "step": 39600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.560023831524358e-05,
+      "loss": 1.9368,
+      "step": 39744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.548867445779015e-05,
+      "loss": 1.9274,
+      "step": 39888
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.53767534001808e-05,
+      "loss": 1.9237,
+      "step": 40032
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.52644762689219e-05,
+      "loss": 1.9127,
+      "step": 40176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.515184419410382e-05,
+      "loss": 1.9343,
+      "step": 40320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.503885830938949e-05,
+      "loss": 1.9175,
+      "step": 40464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.492551975200298e-05,
+      "loss": 1.8913,
+      "step": 40608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.328125,
+      "learning_rate": 8.481182966271815e-05,
+      "loss": 1.9123,
+      "step": 40752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.125,
+      "learning_rate": 8.469778918584701e-05,
+      "loss": 1.8848,
+      "step": 40896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.458339946922833e-05,
+      "loss": 1.9092,
+      "step": 41040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.4468661664216e-05,
+      "loss": 1.9293,
+      "step": 41184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.435357692566758e-05,
+      "loss": 1.8904,
+      "step": 41328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.875,
+      "learning_rate": 8.42381464119325e-05,
+      "loss": 1.9033,
+      "step": 41472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.412237128484047e-05,
+      "loss": 1.8981,
+      "step": 41616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.125,
+      "learning_rate": 8.400625270968987e-05,
+      "loss": 1.9204,
+      "step": 41760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.78125,
+      "learning_rate": 8.388979185523589e-05,
+      "loss": 1.918,
+      "step": 41904
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.377298989367884e-05,
+      "loss": 1.8975,
+      "step": 42048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.365584800065236e-05,
+      "loss": 1.9176,
+      "step": 42192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.353836735521154e-05,
+      "loss": 1.9063,
+      "step": 42336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.342054913982107e-05,
+      "loss": 1.9107,
+      "step": 42480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.330239454034337e-05,
+      "loss": 1.9103,
+      "step": 42624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.318390474602663e-05,
+      "loss": 1.9036,
+      "step": 42768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.306508094949277e-05,
+      "loss": 1.9126,
+      "step": 42912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.294592434672561e-05,
+      "loss": 1.904,
+      "step": 43056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.282643613705867e-05,
+      "loss": 1.9269,
+      "step": 43200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.270661752316316e-05,
+      "loss": 1.923,
+      "step": 43344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.258646971103584e-05,
+      "loss": 1.9142,
+      "step": 43488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.246599390998699e-05,
+      "loss": 1.902,
+      "step": 43632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.125,
+      "learning_rate": 8.234519133262808e-05,
+      "loss": 1.9054,
+      "step": 43776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.222406319485971e-05,
+      "loss": 1.934,
+      "step": 43920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.210261071585927e-05,
+      "loss": 1.9144,
+      "step": 44064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.6875,
+      "learning_rate": 8.198083511806868e-05,
+      "loss": 1.9006,
+      "step": 44208
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.185873762718217e-05,
+      "loss": 1.914,
+      "step": 44352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.173631947213385e-05,
+      "loss": 1.9198,
+      "step": 44496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.125,
+      "learning_rate": 8.161358188508538e-05,
+      "loss": 1.928,
+      "step": 44640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.149052610141357e-05,
+      "loss": 1.9129,
+      "step": 44784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.136715335969792e-05,
+      "loss": 1.9024,
+      "step": 44928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.12434649017082e-05,
+      "loss": 1.9092,
+      "step": 45072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.111946197239188e-05,
+      "loss": 1.9193,
+      "step": 45216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.099514581986171e-05,
+      "loss": 1.9084,
+      "step": 45360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.087051769538304e-05,
+      "loss": 1.9034,
+      "step": 45504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.074557885336127e-05,
+      "loss": 1.9067,
+      "step": 45648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.062033055132927e-05,
+      "loss": 1.9127,
+      "step": 45792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.049477404993464e-05,
+      "loss": 1.9125,
+      "step": 45936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.036891061292709e-05,
+      "loss": 1.9074,
+      "step": 46080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.02427415071457e-05,
+      "loss": 1.9195,
+      "step": 46224
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.011626800250615e-05,
+      "loss": 1.915,
+      "step": 46368
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.998949137198795e-05,
+      "loss": 1.9123,
+      "step": 46512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.986241289162162e-05,
+      "loss": 1.9281,
+      "step": 46656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.97350338404759e-05,
+      "loss": 1.8771,
+      "step": 46800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.960735550064478e-05,
+      "loss": 1.9318,
+      "step": 46944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.947937915723466e-05,
+      "loss": 1.9139,
+      "step": 47088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.125,
+      "learning_rate": 7.935110609835142e-05,
+      "loss": 1.9126,
+      "step": 47232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.125,
+      "learning_rate": 7.922253761508743e-05,
+      "loss": 1.9131,
+      "step": 47376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.125,
+      "learning_rate": 7.909367500150856e-05,
+      "loss": 1.9272,
+      "step": 47520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.89645195546411e-05,
+      "loss": 1.9072,
+      "step": 47664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.88350725744589e-05,
+      "loss": 1.8913,
+      "step": 47808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.870533536386999e-05,
+      "loss": 1.9215,
+      "step": 47952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.857530922870376e-05,
+      "loss": 1.9269,
+      "step": 48096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.844499547769754e-05,
+      "loss": 1.9072,
+      "step": 48240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.831439542248372e-05,
+      "loss": 1.9031,
+      "step": 48384
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.818351037757627e-05,
+      "loss": 1.9148,
+      "step": 48528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.805234166035771e-05,
+      "loss": 1.9211,
+      "step": 48672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.792089059106574e-05,
+      "loss": 1.9307,
+      "step": 48816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.778915849277997e-05,
+      "loss": 1.9168,
+      "step": 48960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.765714669140869e-05,
+      "loss": 1.9243,
+      "step": 49104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.752485651567539e-05,
+      "loss": 1.9008,
+      "step": 49248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.739228929710544e-05,
+      "loss": 1.9126,
+      "step": 49392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.725944637001277e-05,
+      "loss": 1.9219,
+      "step": 49536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.71263290714863e-05,
+      "loss": 1.9154,
+      "step": 49680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.699293874137657e-05,
+      "loss": 1.9111,
+      "step": 49824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.685927672228226e-05,
+      "loss": 1.9028,
+      "step": 49968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.672534435953662e-05,
+      "loss": 1.9026,
+      "step": 50112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.659114300119395e-05,
+      "loss": 1.9071,
+      "step": 50256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.645667399801609e-05,
+      "loss": 1.921,
+      "step": 50400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.632193870345872e-05,
+      "loss": 1.8933,
+      "step": 50544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.618693847365784e-05,
+      "loss": 1.9139,
+      "step": 50688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.125,
+      "learning_rate": 7.605167466741603e-05,
+      "loss": 1.921,
+      "step": 50832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.591614864618883e-05,
+      "loss": 1.9422,
+      "step": 50976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.578036177407109e-05,
+      "loss": 1.8973,
+      "step": 51120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.875,
+      "learning_rate": 7.564431541778302e-05,
+      "loss": 1.9229,
+      "step": 51264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.550801094665673e-05,
+      "loss": 1.8958,
+      "step": 51408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.537144973262228e-05,
+      "loss": 1.9134,
+      "step": 51552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.52346331501938e-05,
+      "loss": 1.8897,
+      "step": 51696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.509756257645591e-05,
+      "loss": 1.9281,
+      "step": 51840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.496023939104952e-05,
+      "loss": 1.9152,
+      "step": 51984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.482266497615827e-05,
+      "loss": 1.9152,
+      "step": 52128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.468484071649439e-05,
+      "loss": 1.9176,
+      "step": 52272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.454676799928488e-05,
+      "loss": 1.9076,
+      "step": 52416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.44084482142575e-05,
+      "loss": 1.8938,
+      "step": 52560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.426988275362679e-05,
+      "loss": 1.9025,
+      "step": 52704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.413107301208009e-05,
+      "loss": 1.9007,
+      "step": 52848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.399202038676342e-05,
+      "loss": 1.9081,
+      "step": 52992
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.38527262772675e-05,
+      "loss": 1.9028,
+      "step": 53136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.371319208561365e-05,
+      "loss": 1.9027,
+      "step": 53280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.357341921623965e-05,
+      "loss": 1.9105,
+      "step": 53424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.343340907598561e-05,
+      "loss": 1.907,
+      "step": 53568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.875,
+      "learning_rate": 7.329316307407981e-05,
+      "loss": 1.9045,
+      "step": 53712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.315268262212451e-05,
+      "loss": 1.9191,
+      "step": 53856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.301196913408182e-05,
+      "loss": 1.9195,
+      "step": 54000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.287102402625931e-05,
+      "loss": 1.881,
+      "step": 54144
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.27298487172959e-05,
+      "loss": 1.918,
+      "step": 54288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.258844462814755e-05,
+      "loss": 1.9015,
+      "step": 54432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.244681318207287e-05,
+      "loss": 1.91,
+      "step": 54576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.23049558046189e-05,
+      "loss": 1.8926,
+      "step": 54720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.216287392360674e-05,
+      "loss": 1.916,
+      "step": 54864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.20205689691171e-05,
+      "loss": 1.9072,
+      "step": 55008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.187804237347603e-05,
+      "loss": 1.895,
+      "step": 55152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.173529557124042e-05,
+      "loss": 1.9027,
+      "step": 55296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.159232999918353e-05,
+      "loss": 1.9182,
+      "step": 55440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.144914709628066e-05,
+      "loss": 1.9177,
+      "step": 55584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.125,
+      "learning_rate": 7.130574830369454e-05,
+      "loss": 1.9248,
+      "step": 55728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.116213506476087e-05,
+      "loss": 1.9145,
+      "step": 55872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.101830882497382e-05,
+      "loss": 1.8965,
+      "step": 56016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.08742710319714e-05,
+      "loss": 1.9228,
+      "step": 56160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.0730023135521e-05,
+      "loss": 1.9277,
+      "step": 56304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.25,
+      "learning_rate": 7.058556658750471e-05,
+      "loss": 1.9116,
+      "step": 56448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.25,
+      "learning_rate": 7.044090284190476e-05,
+      "loss": 1.9069,
+      "step": 56592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.029603335478883e-05,
+      "loss": 1.9087,
+      "step": 56736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.015095958429544e-05,
+      "loss": 1.9116,
+      "step": 56880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.000568299061929e-05,
+      "loss": 1.9132,
+      "step": 57024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.986020503599651e-05,
+      "loss": 1.9259,
+      "step": 57168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.971452718468991e-05,
+      "loss": 1.9268,
+      "step": 57312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.956865090297439e-05,
+      "loss": 1.9186,
+      "step": 57456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.942257765912206e-05,
+      "loss": 1.891,
+      "step": 57600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.927630892338746e-05,
+      "loss": 1.9136,
+      "step": 57744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.912984616799283e-05,
+      "loss": 1.9042,
+      "step": 57888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.898319086711321e-05,
+      "loss": 1.8964,
+      "step": 58032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.88363444968617e-05,
+      "loss": 1.9057,
+      "step": 58176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.868930853527452e-05,
+      "loss": 1.8916,
+      "step": 58320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.854208446229616e-05,
+      "loss": 1.9167,
+      "step": 58464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.839467375976449e-05,
+      "loss": 1.8944,
+      "step": 58608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.824707791139584e-05,
+      "loss": 1.9114,
+      "step": 58752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.809929840277008e-05,
+      "loss": 1.8976,
+      "step": 58896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.795133672131564e-05,
+      "loss": 1.9072,
+      "step": 59040
+    }
+  ],
+  "logging_steps": 144,
+  "max_steps": 147005,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1440,
+  "total_flos": 2.4734817975942513e+20,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e7f668272e1e3c39c7953a047fd5f2806fb8c3611ae881516e8311e33ea8373
+size 4920