Training in progress, step 10

config.json

@@ -18,9 +18,9 @@
   "hidden_act": "silu",
   "hidden_size": 1012,
   "initializer_range": 0.02,
-  "intermediate_size": 4056,
+  "intermediate_size": 3840,
   "kv_lora_rank": 64,
-  "max_position_embeddings": 2048,
+  "max_position_embeddings": 1024,
   "model_type": "deepseek_v3",
   "moe_intermediate_size": 2048,
   "moe_layer_freq": 1,
@@ -30,7 +30,7 @@
   "norm_topk_prob": true,
   "num_attention_heads": 16,
   "num_experts_per_tok": 2,
-  "num_hidden_layers": 20,
+  "num_hidden_layers": 16,
   "num_key_value_heads": 16,
   "num_nextn_predict_layers": 1,
   "pad_token_id": 0,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c926480da0d9471c7b339a0ecd629e0fdd763b453e0ce511bbd2fa82eb53f8a3
-size 4156744752
+oid sha256:a437902138f34a8b05e2cb92387dea74971f269b96080499dc8bff3e97c306c9
+size 3299862016

runs/Oct07_12-38-56_5dc35bd65435/events.out.tfevents.1759840744.5dc35bd65435.47279.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39253b7ac7e3e6a54908511075fbca3b01be996a038ac151b919e60fe3c7b5c0
+size 8205

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb08b65ddf8cfe9535a7296485be97b38a5d3726d26656631c74af48caeb158c
+oid sha256:81cd0fcf4f4aed1a9a04f7cfa765c51e2322bb13c6e8ac2e141b4b299658a277
 size 5649