Upload JambaForCausalLM

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "ai21labs/Jamba-v0.1",
+  "_name_or_path": "isemmanuelolowe/PyJamba1B",
   "architectures": [
     "JambaForCausalLM"
   ],
@@ -18,9 +18,9 @@
   "expert_layer_offset": 1,
   "expert_layer_period": 2,
   "hidden_act": "silu",
-  "hidden_size": 4096,
+  "hidden_size": 1024,
   "initializer_range": 0.02,
-  "intermediate_size": 14336,
+  "intermediate_size": 2048,
   "mamba_conv_bias": true,
   "mamba_d_conv": 4,
   "mamba_d_state": 16,
@@ -31,34 +31,19 @@
   "model_type": "jamba",
   "n_ctx": 262144,
   "num_attention_heads": 32,
-  "num_experts": 16,
+  "num_experts": 32,
   "num_experts_per_tok": 2,
-  "num_hidden_layers": 32,
+  "num_hidden_layers": 8,
   "num_key_value_heads": 8,
   "output_router_logits": false,
   "pad_token_id": 0,
-  "quantization_config": {
-    "_load_in_4bit": true,
-    "_load_in_8bit": false,
-    "bnb_4bit_compute_dtype": "bfloat16",
-    "bnb_4bit_quant_storage": "bfloat16",
-    "bnb_4bit_quant_type": "nf4",
-    "bnb_4bit_use_double_quant": true,
-    "llm_int8_enable_fp32_cpu_offload": false,
-    "llm_int8_has_fp16_weight": false,
-    "llm_int8_skip_modules": null,
-    "llm_int8_threshold": 6.0,
-    "load_in_4bit": true,
-    "load_in_8bit": false,
-    "quant_method": "bitsandbytes"
-  },
   "rms_norm_eps": 1e-06,
   "router_aux_loss_coef": 0.001,
   "sliding_window": null,
   "tie_word_embeddings": false,
-  "torch_dtype": "bfloat16",
+  "torch_dtype": "float32",
   "transformers_version": "4.40.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "use_mamba_kernels": true,
   "vocab_size": 65536
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d7f9f98d09b6238684e825a1d1260e2ceaa0ef18bce9c23a3869039afd07d9a
+oid sha256:57420b913700953ace3354ebb72959a9f249eaaad9cdb72246bf94a6d24e0d19
 size 4078582904