checkpoint-900 for eval

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "cowshed2_base/checkpoint-300",
+  "_name_or_path": "cowshed2_base/checkpoint-900",
   "architectures": [
     "LlamaForCausalLM"
   ],
@@ -8,14 +8,14 @@
   "bos_token_id": 50257,
   "eos_token_id": 50256,
   "hidden_act": "silu",
-  "hidden_size": 1024,
+  "hidden_size": 768,
   "initializer_range": 0.02,
-  "intermediate_size": 2048,
+  "intermediate_size": 4096,
   "max_position_embeddings": 4096,
   "mlp_bias": false,
   "model_type": "llama",
   "num_attention_heads": 16,
-  "num_hidden_layers": 24,
+  "num_hidden_layers": 16,
   "num_key_value_heads": 16,
   "pad_token_id": 50258,
   "pretraining_tp": 1,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cc424a9d4f483f90a0e78e303d8e46446910689919f0ad255530fc2227f481b8
-size 1418948944
+oid sha256:3b6233ff7838ee293dbfe8688910bbe3d6abcced4ed2adaf14ddf7551056a838
+size 1064160256