Upload folder using huggingface_hub

README.md

@@ -6,7 +6,7 @@ base_model: lmzheng/grok-1
 
 # Grok-1-W4A8KV8
 ## Introduction
-  This model was created by applying [Quark](https://quark.docs.amd.com/latest/index.html) with calibration samples from Pile dataset. The model is TP8 sharded.
+  This model was created by applying [Quark](https://quark.docs.amd.com/latest/index.html) with calibration samples from Pile dataset.
 ## Quantization Stragegy
   - ***Quantized Layers***: All linear layers excluding "lm_head", "*.gate"
   - ***Weight***: FP8 symmetric per-tensor, additionally, INT4 symmetric per-channel for MoE linear

config.json

@@ -1,10 +1,9 @@
 {
-  "_name_or_path": "/data/llm/grok-1-sglang-tp1",
+  "_name_or_path": "/group/amdneuralopt/huggingface/pretrained_models/grok-1-sglang-tp1",
   "architectures": [
     "Grok1ModelForCausalLM"
   ],
-  "attn_logit_softcapping": 30.0,
-  "attn_output_multiplier": 1.0,
+  "attn_output_multiplier": 0.08838834764831845,
   "auto_map": {
     "AutoConfig": "configuration_grok1.Grok1Config",
     "AutoModel": "modeling_grok1.Grok1Model",
@@ -13,21 +12,19 @@
   "bos_token_id": 1,
   "embedding_multiplier_scale": 78.38367176906169,
   "eos_token_id": 2,
-  "final_logit_softcapping": 50.0,
-  "head_dim": 128,
   "hidden_size": 6144,
   "intermediate_size": 32768,
-  "max_attn_value": 1.0,
-  "max_position_embeddings": 32768,
+  "max_attn_value": 30.0,
+  "max_position_embeddings": 8192,
   "model_type": "grok-1",
   "num_attention_heads": 48,
   "num_experts": 8,
   "num_experts_per_tok": 2,
   "num_hidden_layers": 64,
   "num_key_value_heads": 8,
-  "num_local_experts": 8,
   "output_multiplier_scale": 0.5773502691896257,
   "output_router_logits": false,
+  "pad_token_id": 0,
   "quantization_config": {
     "activation_scheme": "static",
     "export": {
@@ -115,11 +112,8 @@
       "group": "column"
     }
   },
-  "replicate_lm_head": false,
   "rms_norm_eps": 1e-05,
-  "rope_theta": 10000.0,
   "router_aux_loss_coef": 0.001,
-  "router_logit_softcapping": 30.0,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.47.1",
   "use_cache": true,

generation_config.json

@@ -2,5 +2,6 @@
   "_from_model_config": true,
   "bos_token_id": 1,
   "eos_token_id": 2,
+  "pad_token_id": 0,
   "transformers_version": "4.47.1"
 }