extra2

checkpoint-6341/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.42.3"
+}

checkpoint-6341/latest

@@ -0,0 +1 @@
+global_step6341

checkpoint-6341/model.safetensors.index.json

@@ -0,0 +1,298 @@
+{
+  "metadata": {
+    "total_size": 5403120640
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

checkpoint-6341/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
+  }
+}

checkpoint-6341/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

checkpoint-6341/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": false,
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

checkpoint-6341/trainer_state.json

@@ -0,0 +1,1758 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 200,
+  "global_step": 6341,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004731114966093676,
+      "grad_norm": 0.9585382342338562,
+      "learning_rate": 7.235790156711095e-05,
+      "loss": 1.19,
+      "step": 30
+    },
+    {
+      "epoch": 0.009462229932187352,
+      "grad_norm": 0.9438452124595642,
+      "learning_rate": 8.817139967814685e-05,
+      "loss": 1.0589,
+      "step": 60
+    },
+    {
+      "epoch": 0.014193344898281028,
+      "grad_norm": 0.9442492723464966,
+      "learning_rate": 9.722413360750843e-05,
+      "loss": 1.0764,
+      "step": 90
+    },
+    {
+      "epoch": 0.018924459864374705,
+      "grad_norm": 0.8840267658233643,
+      "learning_rate": 9.994621104255655e-05,
+      "loss": 1.0847,
+      "step": 120
+    },
+    {
+      "epoch": 0.02365557483046838,
+      "grad_norm": 0.8207218050956726,
+      "learning_rate": 9.985445340927068e-05,
+      "loss": 1.0912,
+      "step": 150
+    },
+    {
+      "epoch": 0.028386689796562056,
+      "grad_norm": 0.8883314728736877,
+      "learning_rate": 9.975953171966461e-05,
+      "loss": 1.0608,
+      "step": 180
+    },
+    {
+      "epoch": 0.03154076644062451,
+      "eval_loss": 1.2097724676132202,
+      "eval_runtime": 3.756,
+      "eval_samples_per_second": 26.89,
+      "eval_steps_per_second": 3.461,
+      "step": 200
+    },
+    {
+      "epoch": 0.03311780476265573,
+      "grad_norm": 0.7577874064445496,
+      "learning_rate": 9.966461003005853e-05,
+      "loss": 1.0802,
+      "step": 210
+    },
+    {
+      "epoch": 0.03784891972874941,
+      "grad_norm": 1.4911932945251465,
+      "learning_rate": 9.956968834045246e-05,
+      "loss": 1.0397,
+      "step": 240
+    },
+    {
+      "epoch": 0.04258003469484308,
+      "grad_norm": 0.8236317038536072,
+      "learning_rate": 9.947476665084638e-05,
+      "loss": 1.0575,
+      "step": 270
+    },
+    {
+      "epoch": 0.04731114966093676,
+      "grad_norm": 0.7883521318435669,
+      "learning_rate": 9.937984496124031e-05,
+      "loss": 1.0369,
+      "step": 300
+    },
+    {
+      "epoch": 0.05204226462703044,
+      "grad_norm": 0.7798565626144409,
+      "learning_rate": 9.928492327163424e-05,
+      "loss": 1.0354,
+      "step": 330
+    },
+    {
+      "epoch": 0.05677337959312411,
+      "grad_norm": 0.7784315943717957,
+      "learning_rate": 9.919000158202817e-05,
+      "loss": 1.0341,
+      "step": 360
+    },
+    {
+      "epoch": 0.06150449455921779,
+      "grad_norm": 0.836300790309906,
+      "learning_rate": 9.909507989242209e-05,
+      "loss": 1.0272,
+      "step": 390
+    },
+    {
+      "epoch": 0.06308153288124901,
+      "eval_loss": 1.1889104843139648,
+      "eval_runtime": 3.7553,
+      "eval_samples_per_second": 26.895,
+      "eval_steps_per_second": 3.462,
+      "step": 400
+    },
+    {
+      "epoch": 0.06623560952531146,
+      "grad_norm": 0.7245925664901733,
+      "learning_rate": 9.900015820281602e-05,
+      "loss": 1.0256,
+      "step": 420
+    },
+    {
+      "epoch": 0.07096672449140515,
+      "grad_norm": 0.8321049213409424,
+      "learning_rate": 9.890523651320994e-05,
+      "loss": 1.0332,
+      "step": 450
+    },
+    {
+      "epoch": 0.07569783945749882,
+      "grad_norm": 0.7657173275947571,
+      "learning_rate": 9.881031482360387e-05,
+      "loss": 1.0221,
+      "step": 480
+    },
+    {
+      "epoch": 0.08042895442359249,
+      "grad_norm": 0.7464463114738464,
+      "learning_rate": 9.871539313399779e-05,
+      "loss": 0.9911,
+      "step": 510
+    },
+    {
+      "epoch": 0.08516006938968616,
+      "grad_norm": 0.7290617227554321,
+      "learning_rate": 9.862047144439172e-05,
+      "loss": 1.0258,
+      "step": 540
+    },
+    {
+      "epoch": 0.08989118435577985,
+      "grad_norm": 0.7311350703239441,
+      "learning_rate": 9.852554975478564e-05,
+      "loss": 1.0165,
+      "step": 570
+    },
+    {
+      "epoch": 0.09462229932187352,
+      "grad_norm": 0.8087915182113647,
+      "learning_rate": 9.843062806517957e-05,
+      "loss": 0.9716,
+      "step": 600
+    },
+    {
+      "epoch": 0.09462229932187352,
+      "eval_loss": 1.1471492052078247,
+      "eval_runtime": 3.7536,
+      "eval_samples_per_second": 26.907,
+      "eval_steps_per_second": 3.463,
+      "step": 600
+    },
+    {
+      "epoch": 0.09935341428796719,
+      "grad_norm": 0.7442970275878906,
+      "learning_rate": 9.833570637557348e-05,
+      "loss": 0.9747,
+      "step": 630
+    },
+    {
+      "epoch": 0.10408452925406088,
+      "grad_norm": 0.9510965347290039,
+      "learning_rate": 9.824078468596742e-05,
+      "loss": 0.9582,
+      "step": 660
+    },
+    {
+      "epoch": 0.10881564422015455,
+      "grad_norm": 0.6995567083358765,
+      "learning_rate": 9.814586299636133e-05,
+      "loss": 1.0118,
+      "step": 690
+    },
+    {
+      "epoch": 0.11354675918624822,
+      "grad_norm": 0.9319436550140381,
+      "learning_rate": 9.805094130675526e-05,
+      "loss": 0.9815,
+      "step": 720
+    },
+    {
+      "epoch": 0.11827787415234191,
+      "grad_norm": 0.7033783793449402,
+      "learning_rate": 9.795601961714918e-05,
+      "loss": 0.9738,
+      "step": 750
+    },
+    {
+      "epoch": 0.12300898911843558,
+      "grad_norm": 0.6606217622756958,
+      "learning_rate": 9.786109792754311e-05,
+      "loss": 0.961,
+      "step": 780
+    },
+    {
+      "epoch": 0.12616306576249803,
+      "eval_loss": 1.125948190689087,
+      "eval_runtime": 3.7557,
+      "eval_samples_per_second": 26.892,
+      "eval_steps_per_second": 3.461,
+      "step": 800
+    },
+    {
+      "epoch": 0.12774010408452927,
+      "grad_norm": 0.9087960124015808,
+      "learning_rate": 9.776617623793703e-05,
+      "loss": 0.9734,
+      "step": 810
+    },
+    {
+      "epoch": 0.13247121905062292,
+      "grad_norm": 0.7387025952339172,
+      "learning_rate": 9.767125454833097e-05,
+      "loss": 0.9605,
+      "step": 840
+    },
+    {
+      "epoch": 0.1372023340167166,
+      "grad_norm": 0.7939543724060059,
+      "learning_rate": 9.757633285872489e-05,
+      "loss": 0.952,
+      "step": 870
+    },
+    {
+      "epoch": 0.1419334489828103,
+      "grad_norm": 1.1417864561080933,
+      "learning_rate": 9.748141116911882e-05,
+      "loss": 0.9113,
+      "step": 900
+    },
+    {
+      "epoch": 0.14666456394890395,
+      "grad_norm": 0.7591778635978699,
+      "learning_rate": 9.738648947951274e-05,
+      "loss": 0.9565,
+      "step": 930
+    },
+    {
+      "epoch": 0.15139567891499764,
+      "grad_norm": 0.759545087814331,
+      "learning_rate": 9.729156778990667e-05,
+      "loss": 0.9401,
+      "step": 960
+    },
+    {
+      "epoch": 0.1561267938810913,
+      "grad_norm": 0.700552761554718,
+      "learning_rate": 9.719664610030059e-05,
+      "loss": 0.9447,
+      "step": 990
+    },
+    {
+      "epoch": 0.15770383220312253,
+      "eval_loss": 1.0677810907363892,
+      "eval_runtime": 3.7551,
+      "eval_samples_per_second": 26.897,
+      "eval_steps_per_second": 3.462,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16085790884718498,
+      "grad_norm": 0.6673519015312195,
+      "learning_rate": 9.710172441069452e-05,
+      "loss": 0.8919,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16558902381327867,
+      "grad_norm": 0.8046931028366089,
+      "learning_rate": 9.700680272108844e-05,
+      "loss": 0.9136,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17032013877937233,
+      "grad_norm": 0.7277413606643677,
+      "learning_rate": 9.691188103148237e-05,
+      "loss": 0.9001,
+      "step": 1080
+    },
+    {
+      "epoch": 0.175051253745466,
+      "grad_norm": 0.661359429359436,
+      "learning_rate": 9.681695934187629e-05,
+      "loss": 0.9119,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1797823687115597,
+      "grad_norm": 0.7349006533622742,
+      "learning_rate": 9.672203765227022e-05,
+      "loss": 0.8825,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18451348367765336,
+      "grad_norm": 0.7114729285240173,
+      "learning_rate": 9.662711596266414e-05,
+      "loss": 0.8872,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18924459864374704,
+      "grad_norm": 0.6496574282646179,
+      "learning_rate": 9.653219427305807e-05,
+      "loss": 0.8809,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18924459864374704,
+      "eval_loss": 1.0253973007202148,
+      "eval_runtime": 3.7532,
+      "eval_samples_per_second": 26.91,
+      "eval_steps_per_second": 3.464,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19397571360984073,
+      "grad_norm": 0.6576619744300842,
+      "learning_rate": 9.643727258345198e-05,
+      "loss": 0.876,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19870682857593439,
+      "grad_norm": 0.666749119758606,
+      "learning_rate": 9.634235089384591e-05,
+      "loss": 0.8877,
+      "step": 1260
+    },
+    {
+      "epoch": 0.20343794354202807,
+      "grad_norm": 0.7769750952720642,
+      "learning_rate": 9.624742920423983e-05,
+      "loss": 0.8894,
+      "step": 1290
+    },
+    {
+      "epoch": 0.20816905850812176,
+      "grad_norm": 0.6562801599502563,
+      "learning_rate": 9.615250751463376e-05,
+      "loss": 0.8912,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21290017347421542,
+      "grad_norm": 0.6531364917755127,
+      "learning_rate": 9.605758582502768e-05,
+      "loss": 0.875,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2176312884403091,
+      "grad_norm": 0.6414660811424255,
+      "learning_rate": 9.596266413542163e-05,
+      "loss": 0.8721,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22078536508437155,
+      "eval_loss": 1.0128834247589111,
+      "eval_runtime": 3.7539,
+      "eval_samples_per_second": 26.906,
+      "eval_steps_per_second": 3.463,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2223624034064028,
+      "grad_norm": 0.8413099646568298,
+      "learning_rate": 9.586774244581554e-05,
+      "loss": 0.8807,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22709351837249644,
+      "grad_norm": 0.6748294830322266,
+      "learning_rate": 9.577282075620947e-05,
+      "loss": 0.8245,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23182463333859013,
+      "grad_norm": 0.7067525386810303,
+      "learning_rate": 9.567789906660339e-05,
+      "loss": 0.8767,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23655574830468382,
+      "grad_norm": 1.074791431427002,
+      "learning_rate": 9.558297737699732e-05,
+      "loss": 0.8856,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24128686327077747,
+      "grad_norm": 0.7461240887641907,
+      "learning_rate": 9.548805568739124e-05,
+      "loss": 0.8759,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24601797823687116,
+      "grad_norm": 0.6231616139411926,
+      "learning_rate": 9.539313399778517e-05,
+      "loss": 0.837,
+      "step": 1560
+    },
+    {
+      "epoch": 0.25074909320296485,
+      "grad_norm": 0.7053641080856323,
+      "learning_rate": 9.529821230817909e-05,
+      "loss": 0.8763,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25232613152499606,
+      "eval_loss": 0.9505324959754944,
+      "eval_runtime": 3.7563,
+      "eval_samples_per_second": 26.888,
+      "eval_steps_per_second": 3.461,
+      "step": 1600
+    },
+    {
+      "epoch": 0.25548020816905853,
+      "grad_norm": 0.6484207510948181,
+      "learning_rate": 9.520329061857302e-05,
+      "loss": 0.8787,
+      "step": 1620
+    },
+    {
+      "epoch": 0.26021132313515216,
+      "grad_norm": 0.5929827094078064,
+      "learning_rate": 9.510836892896694e-05,
+      "loss": 0.844,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26494243810124585,
+      "grad_norm": 0.6840829849243164,
+      "learning_rate": 9.501344723936087e-05,
+      "loss": 0.8492,
+      "step": 1680
+    },
+    {
+      "epoch": 0.26967355306733953,
+      "grad_norm": 0.7365448474884033,
+      "learning_rate": 9.491852554975479e-05,
+      "loss": 0.8584,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2744046680334332,
+      "grad_norm": 0.6528182029724121,
+      "learning_rate": 9.482360386014872e-05,
+      "loss": 0.8346,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2791357829995269,
+      "grad_norm": 0.6200223565101624,
+      "learning_rate": 9.472868217054263e-05,
+      "loss": 0.8008,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2838668979656206,
+      "grad_norm": 0.7503982186317444,
+      "learning_rate": 9.463376048093657e-05,
+      "loss": 0.8197,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2838668979656206,
+      "eval_loss": 0.9286572933197021,
+      "eval_runtime": 3.7535,
+      "eval_samples_per_second": 26.908,
+      "eval_steps_per_second": 3.463,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2885980129317142,
+      "grad_norm": 0.6671140193939209,
+      "learning_rate": 9.453883879133048e-05,
+      "loss": 0.8405,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2933291278978079,
+      "grad_norm": 0.7057023048400879,
+      "learning_rate": 9.444391710172441e-05,
+      "loss": 0.7822,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2980602428639016,
+      "grad_norm": 0.8120527267456055,
+      "learning_rate": 9.434899541211833e-05,
+      "loss": 0.8416,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3027913578299953,
+      "grad_norm": 0.622718334197998,
+      "learning_rate": 9.425407372251228e-05,
+      "loss": 0.8174,
+      "step": 1920
+    },
+    {
+      "epoch": 0.30752247279608896,
+      "grad_norm": 0.6605896353721619,
+      "learning_rate": 9.41591520329062e-05,
+      "loss": 0.8003,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3122535877621826,
+      "grad_norm": 0.7473495006561279,
+      "learning_rate": 9.406423034330012e-05,
+      "loss": 0.798,
+      "step": 1980
+    },
+    {
+      "epoch": 0.31540766440624507,
+      "eval_loss": 0.8976284861564636,
+      "eval_runtime": 3.7537,
+      "eval_samples_per_second": 26.907,
+      "eval_steps_per_second": 3.463,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3169847027282763,
+      "grad_norm": 0.7177520394325256,
+      "learning_rate": 9.396930865369404e-05,
+      "loss": 0.8168,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32171581769436997,
+      "grad_norm": 0.7600869536399841,
+      "learning_rate": 9.387438696408797e-05,
+      "loss": 0.7918,
+      "step": 2040
+    },
+    {
+      "epoch": 0.32644693266046365,
+      "grad_norm": 0.7001503109931946,
+      "learning_rate": 9.377946527448189e-05,
+      "loss": 0.7906,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33117804762655734,
+      "grad_norm": 0.6279382705688477,
+      "learning_rate": 9.368454358487582e-05,
+      "loss": 0.7624,
+      "step": 2100
+    },
+    {
+      "epoch": 0.335909162592651,
+      "grad_norm": 0.7481889128684998,
+      "learning_rate": 9.358962189526974e-05,
+      "loss": 0.7849,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34064027755874465,
+      "grad_norm": 0.6797828078269958,
+      "learning_rate": 9.349470020566367e-05,
+      "loss": 0.7899,
+      "step": 2160
+    },
+    {
+      "epoch": 0.34537139252483834,
+      "grad_norm": 0.6929941177368164,
+      "learning_rate": 9.339977851605759e-05,
+      "loss": 0.7703,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3469484308468696,
+      "eval_loss": 0.8858568072319031,
+      "eval_runtime": 3.7538,
+      "eval_samples_per_second": 26.906,
+      "eval_steps_per_second": 3.463,
+      "step": 2200
+    },
+    {
+      "epoch": 0.350102507490932,
+      "grad_norm": 0.698906660079956,
+      "learning_rate": 9.330485682645152e-05,
+      "loss": 0.7724,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3548336224570257,
+      "grad_norm": 0.779211163520813,
+      "learning_rate": 9.320993513684544e-05,
+      "loss": 0.7875,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3595647374231194,
+      "grad_norm": 0.7313475608825684,
+      "learning_rate": 9.311817750355957e-05,
+      "loss": 0.794,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3642958523892131,
+      "grad_norm": 0.6143506169319153,
+      "learning_rate": 9.30232558139535e-05,
+      "loss": 0.7742,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3690269673553067,
+      "grad_norm": 0.6775010824203491,
+      "learning_rate": 9.292833412434741e-05,
+      "loss": 0.7822,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3737580823214004,
+      "grad_norm": 0.7151722311973572,
+      "learning_rate": 9.283341243474134e-05,
+      "loss": 0.7617,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3784891972874941,
+      "grad_norm": 0.6855128407478333,
+      "learning_rate": 9.273849074513526e-05,
+      "loss": 0.7668,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3784891972874941,
+      "eval_loss": 0.8862702250480652,
+      "eval_runtime": 3.7541,
+      "eval_samples_per_second": 26.904,
+      "eval_steps_per_second": 3.463,
+      "step": 2400
+    },
+    {
+      "epoch": 0.38322031225358777,
+      "grad_norm": 0.743325412273407,
+      "learning_rate": 9.26435690555292e-05,
+      "loss": 0.7885,
+      "step": 2430
+    },
+    {
+      "epoch": 0.38795142721968146,
+      "grad_norm": 0.6186659932136536,
+      "learning_rate": 9.254864736592311e-05,
+      "loss": 0.7619,
+      "step": 2460
+    },
+    {
+      "epoch": 0.39268254218577514,
+      "grad_norm": 0.6791619062423706,
+      "learning_rate": 9.245372567631704e-05,
+      "loss": 0.8084,
+      "step": 2490
+    },
+    {
+      "epoch": 0.39741365715186877,
+      "grad_norm": 0.6537867784500122,
+      "learning_rate": 9.235880398671097e-05,
+      "loss": 0.7641,
+      "step": 2520
+    },
+    {
+      "epoch": 0.40214477211796246,
+      "grad_norm": 0.6688680052757263,
+      "learning_rate": 9.22638822971049e-05,
+      "loss": 0.7634,
+      "step": 2550
+    },
+    {
+      "epoch": 0.40687588708405614,
+      "grad_norm": 0.6369423866271973,
+      "learning_rate": 9.216896060749882e-05,
+      "loss": 0.7407,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4100299637281186,
+      "eval_loss": 0.8817442059516907,
+      "eval_runtime": 3.7541,
+      "eval_samples_per_second": 26.904,
+      "eval_steps_per_second": 3.463,
+      "step": 2600
+    },
+    {
+      "epoch": 0.41160700205014983,
+      "grad_norm": 0.6841573119163513,
+      "learning_rate": 9.207403891789275e-05,
+      "loss": 0.7572,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4163381170162435,
+      "grad_norm": 0.625957727432251,
+      "learning_rate": 9.197911722828667e-05,
+      "loss": 0.7493,
+      "step": 2640
+    },
+    {
+      "epoch": 0.42106923198233714,
+      "grad_norm": 0.7467941641807556,
+      "learning_rate": 9.18841955386806e-05,
+      "loss": 0.7468,
+      "step": 2670
+    },
+    {
+      "epoch": 0.42580034694843083,
+      "grad_norm": 0.6891815662384033,
+      "learning_rate": 9.178927384907452e-05,
+      "loss": 0.7698,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4305314619145245,
+      "grad_norm": 0.6197889447212219,
+      "learning_rate": 9.169435215946845e-05,
+      "loss": 0.7588,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4352625768806182,
+      "grad_norm": 0.7140328884124756,
+      "learning_rate": 9.159943046986237e-05,
+      "loss": 0.7569,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4399936918467119,
+      "grad_norm": 0.7718496322631836,
+      "learning_rate": 9.15045087802563e-05,
+      "loss": 0.7448,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4415707301687431,
+      "eval_loss": 0.8855557441711426,
+      "eval_runtime": 3.7544,
+      "eval_samples_per_second": 26.902,
+      "eval_steps_per_second": 3.463,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4447248068128056,
+      "grad_norm": 0.6447039246559143,
+      "learning_rate": 9.140958709065022e-05,
+      "loss": 0.7623,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4494559217788992,
+      "grad_norm": 0.6694769859313965,
+      "learning_rate": 9.131466540104415e-05,
+      "loss": 0.7081,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4541870367449929,
+      "grad_norm": 0.6863081455230713,
+      "learning_rate": 9.121974371143806e-05,
+      "loss": 0.7228,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4589181517110866,
+      "grad_norm": 0.7198454737663269,
+      "learning_rate": 9.1124822021832e-05,
+      "loss": 0.7356,
+      "step": 2910
+    },
+    {
+      "epoch": 0.46364926667718026,
+      "grad_norm": 0.6542885303497314,
+      "learning_rate": 9.102990033222591e-05,
+      "loss": 0.7606,
+      "step": 2940
+    },
+    {
+      "epoch": 0.46838038164327395,
+      "grad_norm": 0.657539963722229,
+      "learning_rate": 9.093497864261984e-05,
+      "loss": 0.7255,
+      "step": 2970
+    },
+    {
+      "epoch": 0.47311149660936763,
+      "grad_norm": 0.819503664970398,
+      "learning_rate": 9.084005695301376e-05,
+      "loss": 0.7184,
+      "step": 3000
+    },
+    {
+      "epoch": 0.47311149660936763,
+      "eval_loss": 0.8140414357185364,
+      "eval_runtime": 3.7531,
+      "eval_samples_per_second": 26.911,
+      "eval_steps_per_second": 3.464,
+      "step": 3000
+    },
+    {
+      "epoch": 0.47784261157546126,
+      "grad_norm": 0.7199704647064209,
+      "learning_rate": 9.074513526340769e-05,
+      "loss": 0.7227,
+      "step": 3030
+    },
+    {
+      "epoch": 0.48257372654155495,
+      "grad_norm": 0.7655025720596313,
+      "learning_rate": 9.065021357380162e-05,
+      "loss": 0.7217,
+      "step": 3060
+    },
+    {
+      "epoch": 0.48730484150764863,
+      "grad_norm": 0.7312873601913452,
+      "learning_rate": 9.055845594051574e-05,
+      "loss": 0.7059,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4920359564737423,
+      "grad_norm": 0.5961809158325195,
+      "learning_rate": 9.046353425090967e-05,
+      "loss": 0.7033,
+      "step": 3120
+    },
+    {
+      "epoch": 0.496767071439836,
+      "grad_norm": 0.6955564022064209,
+      "learning_rate": 9.03686125613036e-05,
+      "loss": 0.7289,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5014981864059297,
+      "grad_norm": 0.6622660160064697,
+      "learning_rate": 9.027369087169752e-05,
+      "loss": 0.6935,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5046522630499921,
+      "eval_loss": 0.7775673270225525,
+      "eval_runtime": 3.754,
+      "eval_samples_per_second": 26.904,
+      "eval_steps_per_second": 3.463,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5062293013720234,
+      "grad_norm": 0.7262014746665955,
+      "learning_rate": 9.017876918209145e-05,
+      "loss": 0.6906,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5109604163381171,
+      "grad_norm": 0.7221697568893433,
+      "learning_rate": 9.008384749248537e-05,
+      "loss": 0.7079,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5156915313042106,
+      "grad_norm": 0.7115603089332581,
+      "learning_rate": 8.99889258028793e-05,
+      "loss": 0.7191,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5204226462703043,
+      "grad_norm": 0.7292232513427734,
+      "learning_rate": 8.989400411327322e-05,
+      "loss": 0.6702,
+      "step": 3300
+    },
+    {
+      "epoch": 0.525153761236398,
+      "grad_norm": 0.741580605506897,
+      "learning_rate": 8.979908242366715e-05,
+      "loss": 0.6762,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5298848762024917,
+      "grad_norm": 0.7870708107948303,
+      "learning_rate": 8.970416073406108e-05,
+      "loss": 0.6838,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5346159911685854,
+      "grad_norm": 0.71812903881073,
+      "learning_rate": 8.9609239044455e-05,
+      "loss": 0.7174,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5361930294906166,
+      "eval_loss": 0.7375061511993408,
+      "eval_runtime": 3.7548,
+      "eval_samples_per_second": 26.899,
+      "eval_steps_per_second": 3.462,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5393471061346791,
+      "grad_norm": 0.7266995906829834,
+      "learning_rate": 8.951431735484893e-05,
+      "loss": 0.6763,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5440782211007728,
+      "grad_norm": 0.7786857485771179,
+      "learning_rate": 8.941939566524284e-05,
+      "loss": 0.7149,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5488093360668664,
+      "grad_norm": 0.7807109355926514,
+      "learning_rate": 8.932447397563677e-05,
+      "loss": 0.6534,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5535404510329601,
+      "grad_norm": 0.6960239410400391,
+      "learning_rate": 8.922955228603069e-05,
+      "loss": 0.7313,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5582715659990538,
+      "grad_norm": 0.586615264415741,
+      "learning_rate": 8.913463059642462e-05,
+      "loss": 0.6579,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5630026809651475,
+      "grad_norm": 0.9740248918533325,
+      "learning_rate": 8.903970890681854e-05,
+      "loss": 0.7013,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5677337959312412,
+      "grad_norm": 0.6628558039665222,
+      "learning_rate": 8.894478721721247e-05,
+      "loss": 0.6546,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5677337959312412,
+      "eval_loss": 0.7031014561653137,
+      "eval_runtime": 3.7542,
+      "eval_samples_per_second": 26.903,
+      "eval_steps_per_second": 3.463,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5724649108973348,
+      "grad_norm": 0.6030669808387756,
+      "learning_rate": 8.884986552760639e-05,
+      "loss": 0.7146,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5771960258634284,
+      "grad_norm": 0.6010313034057617,
+      "learning_rate": 8.875494383800032e-05,
+      "loss": 0.6816,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5819271408295221,
+      "grad_norm": 0.6319311857223511,
+      "learning_rate": 8.866002214839425e-05,
+      "loss": 0.6642,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5866582557956158,
+      "grad_norm": 0.6059941053390503,
+      "learning_rate": 8.856510045878817e-05,
+      "loss": 0.6998,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5913893707617095,
+      "grad_norm": 0.5976997017860413,
+      "learning_rate": 8.84701787691821e-05,
+      "loss": 0.6694,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5961204857278032,
+      "grad_norm": 0.6985177993774414,
+      "learning_rate": 8.837525707957602e-05,
+      "loss": 0.6402,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5992745623718656,
+      "eval_loss": 0.6977850198745728,
+      "eval_runtime": 3.7545,
+      "eval_samples_per_second": 26.901,
+      "eval_steps_per_second": 3.462,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6008516006938969,
+      "grad_norm": 0.7076742053031921,
+      "learning_rate": 8.828033538996995e-05,
+      "loss": 0.6749,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6055827156599906,
+      "grad_norm": 0.9254401326179504,
+      "learning_rate": 8.818541370036387e-05,
+      "loss": 0.6481,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6103138306260842,
+      "grad_norm": 0.7403334379196167,
+      "learning_rate": 8.80904920107578e-05,
+      "loss": 0.6704,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6150449455921779,
+      "grad_norm": 0.6302973628044128,
+      "learning_rate": 8.799557032115171e-05,
+      "loss": 0.6717,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6197760605582716,
+      "grad_norm": 0.7587308287620544,
+      "learning_rate": 8.790064863154565e-05,
+      "loss": 0.6526,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6245071755243652,
+      "grad_norm": 0.768151581287384,
+      "learning_rate": 8.780572694193956e-05,
+      "loss": 0.6614,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6292382904904589,
+      "grad_norm": 0.662624716758728,
+      "learning_rate": 8.77108052523335e-05,
+      "loss": 0.6471,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6308153288124901,
+      "eval_loss": 0.6685364246368408,
+      "eval_runtime": 3.7533,
+      "eval_samples_per_second": 26.909,
+      "eval_steps_per_second": 3.464,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6339694054565526,
+      "grad_norm": 0.614434540271759,
+      "learning_rate": 8.761588356272743e-05,
+      "loss": 0.6305,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6387005204226462,
+      "grad_norm": 0.7292618751525879,
+      "learning_rate": 8.752096187312134e-05,
+      "loss": 0.632,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6434316353887399,
+      "grad_norm": 0.5890663862228394,
+      "learning_rate": 8.742604018351527e-05,
+      "loss": 0.6594,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6481627503548336,
+      "grad_norm": 0.6511669158935547,
+      "learning_rate": 8.733111849390919e-05,
+      "loss": 0.6417,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6528938653209273,
+      "grad_norm": 0.6794877648353577,
+      "learning_rate": 8.723619680430312e-05,
+      "loss": 0.6472,
+      "step": 4140
+    },
+    {
+      "epoch": 0.657624980287021,
+      "grad_norm": 0.5826547145843506,
+      "learning_rate": 8.714127511469704e-05,
+      "loss": 0.6255,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6623560952531147,
+      "grad_norm": 0.8411812782287598,
+      "learning_rate": 8.704635342509097e-05,
+      "loss": 0.6368,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6623560952531147,
+      "eval_loss": 0.6538847088813782,
+      "eval_runtime": 3.7543,
+      "eval_samples_per_second": 26.903,
+      "eval_steps_per_second": 3.463,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6670872102192084,
+      "grad_norm": 0.5682166218757629,
+      "learning_rate": 8.69514317354849e-05,
+      "loss": 0.6269,
+      "step": 4230
+    },
+    {
+      "epoch": 0.671818325185302,
+      "grad_norm": 0.6340855360031128,
+      "learning_rate": 8.685651004587882e-05,
+      "loss": 0.6423,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6765494401513957,
+      "grad_norm": 0.6693681478500366,
+      "learning_rate": 8.676158835627275e-05,
+      "loss": 0.6471,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6812805551174893,
+      "grad_norm": 0.6101056337356567,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.6168,
+      "step": 4320
+    },
+    {
+      "epoch": 0.686011670083583,
+      "grad_norm": 0.6096228361129761,
+      "learning_rate": 8.65717449770606e-05,
+      "loss": 0.6494,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6907427850496767,
+      "grad_norm": 0.6632306575775146,
+      "learning_rate": 8.647682328745452e-05,
+      "loss": 0.664,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6938968616937392,
+      "eval_loss": 0.6377571225166321,
+      "eval_runtime": 3.756,
+      "eval_samples_per_second": 26.89,
+      "eval_steps_per_second": 3.461,
+      "step": 4400
+    },
+    {
+      "epoch": 0.6954739000157704,
+      "grad_norm": 0.6547721028327942,
+      "learning_rate": 8.638190159784845e-05,
+      "loss": 0.6091,
+      "step": 4410
+    },
+    {
+      "epoch": 0.700205014981864,
+      "grad_norm": 0.6063847541809082,
+      "learning_rate": 8.628697990824237e-05,
+      "loss": 0.6055,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7049361299479577,
+      "grad_norm": 0.6687933802604675,
+      "learning_rate": 8.61920582186363e-05,
+      "loss": 0.601,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7096672449140514,
+      "grad_norm": 0.701770007610321,
+      "learning_rate": 8.609713652903021e-05,
+      "loss": 0.6064,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7143983598801451,
+      "grad_norm": 0.6652805209159851,
+      "learning_rate": 8.600221483942414e-05,
+      "loss": 0.653,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7191294748462388,
+      "grad_norm": 0.6469018459320068,
+      "learning_rate": 8.590729314981806e-05,
+      "loss": 0.6019,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7238605898123325,
+      "grad_norm": 0.6343564391136169,
+      "learning_rate": 8.5812371460212e-05,
+      "loss": 0.6083,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7254376281343636,
+      "eval_loss": 0.6411118507385254,
+      "eval_runtime": 3.754,
+      "eval_samples_per_second": 26.905,
+      "eval_steps_per_second": 3.463,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7285917047784262,
+      "grad_norm": 0.5817134976387024,
+      "learning_rate": 8.571744977060592e-05,
+      "loss": 0.602,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7333228197445197,
+      "grad_norm": 0.5552039742469788,
+      "learning_rate": 8.562252808099984e-05,
+      "loss": 0.6223,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7380539347106134,
+      "grad_norm": 0.6455065011978149,
+      "learning_rate": 8.552760639139377e-05,
+      "loss": 0.5865,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7427850496767071,
+      "grad_norm": 0.6448588371276855,
+      "learning_rate": 8.543268470178769e-05,
+      "loss": 0.6126,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7475161646428008,
+      "grad_norm": 0.6447100639343262,
+      "learning_rate": 8.533776301218162e-05,
+      "loss": 0.6167,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7522472796088945,
+      "grad_norm": 0.6894412636756897,
+      "learning_rate": 8.524284132257555e-05,
+      "loss": 0.5851,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7569783945749882,
+      "grad_norm": 0.6036236882209778,
+      "learning_rate": 8.514791963296947e-05,
+      "loss": 0.6025,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7569783945749882,
+      "eval_loss": 0.6117845177650452,
+      "eval_runtime": 3.7554,
+      "eval_samples_per_second": 26.894,
+      "eval_steps_per_second": 3.462,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7617095095410819,
+      "grad_norm": 0.6214340925216675,
+      "learning_rate": 8.50529979433634e-05,
+      "loss": 0.6145,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7664406245071755,
+      "grad_norm": 0.6933445334434509,
+      "learning_rate": 8.495807625375732e-05,
+      "loss": 0.6184,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7711717394732692,
+      "grad_norm": 0.5649739503860474,
+      "learning_rate": 8.486315456415125e-05,
+      "loss": 0.5996,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7759028544393629,
+      "grad_norm": 0.6250168085098267,
+      "learning_rate": 8.476823287454517e-05,
+      "loss": 0.5762,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7806339694054566,
+      "grad_norm": 1.7125053405761719,
+      "learning_rate": 8.46733111849391e-05,
+      "loss": 0.5716,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7853650843715503,
+      "grad_norm": 0.5721966028213501,
+      "learning_rate": 8.457838949533302e-05,
+      "loss": 0.5612,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7885191610156127,
+      "eval_loss": 0.5980841517448425,
+      "eval_runtime": 3.7547,
+      "eval_samples_per_second": 26.9,
+      "eval_steps_per_second": 3.462,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7900961993376439,
+      "grad_norm": 0.6716078519821167,
+      "learning_rate": 8.448346780572695e-05,
+      "loss": 0.5765,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7948273143037375,
+      "grad_norm": 0.6005885601043701,
+      "learning_rate": 8.438854611612086e-05,
+      "loss": 0.5941,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7995584292698312,
+      "grad_norm": 0.6507188081741333,
+      "learning_rate": 8.42936244265148e-05,
+      "loss": 0.5827,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8042895442359249,
+      "grad_norm": 0.7276827096939087,
+      "learning_rate": 8.419870273690871e-05,
+      "loss": 0.5555,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8090206592020186,
+      "grad_norm": 0.6792399287223816,
+      "learning_rate": 8.410378104730264e-05,
+      "loss": 0.5724,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8137517741681123,
+      "grad_norm": 0.7074045538902283,
+      "learning_rate": 8.400885935769656e-05,
+      "loss": 0.5724,
+      "step": 5160
+    },
+    {
+      "epoch": 0.818482889134206,
+      "grad_norm": 0.6056311130523682,
+      "learning_rate": 8.391393766809049e-05,
+      "loss": 0.5546,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8200599274562372,
+      "eval_loss": 0.5805890560150146,
+      "eval_runtime": 3.7561,
+      "eval_samples_per_second": 26.889,
+      "eval_steps_per_second": 3.461,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8232140041002997,
+      "grad_norm": 0.6667674779891968,
+      "learning_rate": 8.381901597848441e-05,
+      "loss": 0.6173,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8279451190663933,
+      "grad_norm": 0.607284426689148,
+      "learning_rate": 8.372409428887834e-05,
+      "loss": 0.5781,
+      "step": 5250
+    },
+    {
+      "epoch": 0.832676234032487,
+      "grad_norm": 0.6476745009422302,
+      "learning_rate": 8.362917259927227e-05,
+      "loss": 0.5667,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8374073489985807,
+      "grad_norm": 0.6668260097503662,
+      "learning_rate": 8.35342509096662e-05,
+      "loss": 0.5456,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8421384639646743,
+      "grad_norm": 0.585110068321228,
+      "learning_rate": 8.343932922006012e-05,
+      "loss": 0.5648,
+      "step": 5340
+    },
+    {
+      "epoch": 0.846869578930768,
+      "grad_norm": 0.6268571019172668,
+      "learning_rate": 8.334757158677425e-05,
+      "loss": 0.555,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8516006938968617,
+      "grad_norm": 0.6197232604026794,
+      "learning_rate": 8.325264989716818e-05,
+      "loss": 0.5333,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8516006938968617,
+      "eval_loss": 0.5601951479911804,
+      "eval_runtime": 3.7534,
+      "eval_samples_per_second": 26.909,
+      "eval_steps_per_second": 3.464,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8563318088629553,
+      "grad_norm": 0.63880455493927,
+      "learning_rate": 8.31577282075621e-05,
+      "loss": 0.5602,
+      "step": 5430
+    },
+    {
+      "epoch": 0.861062923829049,
+      "grad_norm": 0.6235695481300354,
+      "learning_rate": 8.306280651795603e-05,
+      "loss": 0.5604,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8657940387951427,
+      "grad_norm": 0.9000911712646484,
+      "learning_rate": 8.296788482834995e-05,
+      "loss": 0.5654,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8705251537612364,
+      "grad_norm": 0.6557802557945251,
+      "learning_rate": 8.287612719506408e-05,
+      "loss": 0.5962,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8752562687273301,
+      "grad_norm": 0.6231096982955933,
+      "learning_rate": 8.278120550545801e-05,
+      "loss": 0.5636,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8799873836934238,
+      "grad_norm": 0.5984258651733398,
+      "learning_rate": 8.268628381585192e-05,
+      "loss": 0.5616,
+      "step": 5580
+    },
+    {
+      "epoch": 0.8831414603374862,
+      "eval_loss": 0.5611711740493774,
+      "eval_runtime": 3.7542,
+      "eval_samples_per_second": 26.903,
+      "eval_steps_per_second": 3.463,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8847184986595175,
+      "grad_norm": 0.5818042159080505,
+      "learning_rate": 8.259452618256605e-05,
+      "loss": 0.5316,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8894496136256111,
+      "grad_norm": 0.7120912671089172,
+      "learning_rate": 8.249960449295998e-05,
+      "loss": 0.5556,
+      "step": 5640
+    },
+    {
+      "epoch": 0.8941807285917048,
+      "grad_norm": 0.6223446130752563,
+      "learning_rate": 8.24046828033539e-05,
+      "loss": 0.5452,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8989118435577984,
+      "grad_norm": 0.6196858286857605,
+      "learning_rate": 8.230976111374783e-05,
+      "loss": 0.5601,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9036429585238921,
+      "grad_norm": 0.6353973150253296,
+      "learning_rate": 8.221483942414175e-05,
+      "loss": 0.5402,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9083740734899858,
+      "grad_norm": 0.6631510257720947,
+      "learning_rate": 8.211991773453568e-05,
+      "loss": 0.5382,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9131051884560795,
+      "grad_norm": 0.6404465436935425,
+      "learning_rate": 8.20249960449296e-05,
+      "loss": 0.5298,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9146822267781107,
+      "eval_loss": 0.560188353061676,
+      "eval_runtime": 3.7541,
+      "eval_samples_per_second": 26.904,
+      "eval_steps_per_second": 3.463,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9178363034221731,
+      "grad_norm": 0.6810153126716614,
+      "learning_rate": 8.193007435532353e-05,
+      "loss": 0.5159,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9225674183882668,
+      "grad_norm": 0.5828801989555359,
+      "learning_rate": 8.183515266571745e-05,
+      "loss": 0.5155,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9272985333543605,
+      "grad_norm": 0.538987934589386,
+      "learning_rate": 8.174023097611138e-05,
+      "loss": 0.5273,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9320296483204542,
+      "grad_norm": 0.6222363114356995,
+      "learning_rate": 8.16453092865053e-05,
+      "loss": 0.526,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9367607632865479,
+      "grad_norm": 0.542966902256012,
+      "learning_rate": 8.155038759689923e-05,
+      "loss": 0.5653,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9414918782526416,
+      "grad_norm": 0.7064533829689026,
+      "learning_rate": 8.145546590729315e-05,
+      "loss": 0.5207,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9462229932187353,
+      "grad_norm": 0.6652514934539795,
+      "learning_rate": 8.136054421768708e-05,
+      "loss": 0.5342,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9462229932187353,
+      "eval_loss": 0.5476773977279663,
+      "eval_runtime": 3.7543,
+      "eval_samples_per_second": 26.902,
+      "eval_steps_per_second": 3.463,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9509541081848288,
+      "grad_norm": 0.6436010003089905,
+      "learning_rate": 8.126562252808101e-05,
+      "loss": 0.536,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9556852231509225,
+      "grad_norm": 0.5532657504081726,
+      "learning_rate": 8.117070083847494e-05,
+      "loss": 0.5261,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9604163381170162,
+      "grad_norm": 0.6539950370788574,
+      "learning_rate": 8.107577914886886e-05,
+      "loss": 0.5226,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9651474530831099,
+      "grad_norm": 0.5767289996147156,
+      "learning_rate": 8.098085745926279e-05,
+      "loss": 0.534,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9698785680492036,
+      "grad_norm": 0.6355389356613159,
+      "learning_rate": 8.08859357696567e-05,
+      "loss": 0.5282,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9746096830152973,
+      "grad_norm": 0.6711322665214539,
+      "learning_rate": 8.079101408005064e-05,
+      "loss": 0.5384,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9777637596593597,
+      "eval_loss": 0.5372142195701599,
+      "eval_runtime": 3.7547,
+      "eval_samples_per_second": 26.899,
+      "eval_steps_per_second": 3.462,
+      "step": 6200
+    },
+    {
+      "epoch": 0.979340797981391,
+      "grad_norm": 0.5990795493125916,
+      "learning_rate": 8.069609239044455e-05,
+      "loss": 0.4624,
+      "step": 6210
+    },
+    {
+      "epoch": 0.9840719129474846,
+      "grad_norm": 0.6971167325973511,
+      "learning_rate": 8.060117070083848e-05,
+      "loss": 0.5015,
+      "step": 6240
+    },
+    {
+      "epoch": 0.9888030279135783,
+      "grad_norm": 0.6699081659317017,
+      "learning_rate": 8.05062490112324e-05,
+      "loss": 0.5325,
+      "step": 6270
+    },
+    {
+      "epoch": 0.993534142879672,
+      "grad_norm": 0.6347541213035583,
+      "learning_rate": 8.041132732162633e-05,
+      "loss": 0.5255,
+      "step": 6300
+    },
+    {
+      "epoch": 0.9982652578457657,
+      "grad_norm": 0.7587487101554871,
+      "learning_rate": 8.031640563202025e-05,
+      "loss": 0.5154,
+      "step": 6330
+    }
+  ],
+  "logging_steps": 30,
+  "max_steps": 31705,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.9843838888449147e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-6341/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a54e267381bb5495a94c0e010aacd60a26fbc7a49b6188c7488c38be0ea28c37
+size 6267

checkpoint-6341/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)