StarCoder_15b_r_v2_checkpoint_100

README.md

@@ -1,49 +1,33 @@
 ---
 license: bigscience-openrail-m
-datasets:
-- nuprl/MultiPL-T
-metrics:
-- code_eval
 library_name: transformers
 tags:
 - code
 - gpt_bigcode
+datasets:
+- nuprl/MultiPL-T
+metrics:
+- code_eval
 model-index:
-- name: MultiPLCoder-15b-Lua
+- name: MultiPLCoder-15b-OCaml
   results:
   - task:
       type: text-generation
     dataset:
-      type: nuprl/MultiPL-E
       name: MultiPL-HumanEval (Lua)
+      type: nuprl/MultiPL-E
     metrics:
-    - name: pass@1
-      type: pass@1
+    - type: pass@1
       value: 0.31
+      name: pass@1
       verified: true
-- name: MultiPLCoder-15b-Racket
-  results:
-  - task:
-      type: text-generation
-    dataset:
-      type: nuprl/MultiPL-E
-      name: MultiPL-HumanEval (Racket)
-    metrics:
-    - name: pass@1
-      type: pass@1
+    - type: pass@1
       value: 0.21
+      name: pass@1
       verified: true
-- name: MultiPLCoder-15b-OCaml
-  results:
-  - task:
-      type: text-generation
-    dataset:
-      type: nuprl/MultiPL-E
-      name: MultiPL-HumanEval (OCaml)
-    metrics:
-    - name: pass@1
-      type: pass@1
+    - type: pass@1
       value: 0.199
+      name: pass@1
       verified: true
 ---
 # MultiPLCoder-15b

config.json

@@ -1,11 +1,10 @@
 {
-  "_name_or_path": "model_starcoder15b_multiplt_jl_mixed/checkpoint-510",
+  "_name_or_path": "/scratch/a.guha/MutiPL-T/starcoder15b-r-v2/output/checkpoint-100",
   "activation_function": "gelu",
   "architectures": [
     "GPTBigCodeForCausalLM"
   ],
   "attention_softmax_in_fp32": true,
-  "attention_window_size": null,
   "attn_pdrop": 0.1,
   "bos_token_id": 0,
   "embd_pdrop": 0.1,
@@ -25,7 +24,6 @@
   "pad_key_length": true,
   "pre_allocate_kv_cache": false,
   "resid_pdrop": 0.1,
-  "rotary_embedding_scale": -9.210340371976184,
   "scale_attention_softmax_in_fp32": true,
   "scale_attn_weights": true,
   "summary_activation": null,
@@ -34,10 +32,8 @@
   "summary_type": "cls_index",
   "summary_use_proj": true,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.34.1",
-  "use_cache": false,
-  "use_position_embeddings": true,
-  "use_rotary_embeddings": false,
+  "transformers_version": "4.39.1",
+  "use_cache": true,
   "validate_runner_input": true,
   "vocab_size": 49152
 }

generation_config.json

@@ -2,5 +2,5 @@
   "_from_model_config": true,
   "bos_token_id": 0,
   "eos_token_id": 0,
-  "transformers_version": "4.34.1"
+  "transformers_version": "4.39.1"
 }

model-00001-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9e9cff7b5beb5fcde7e0deb31dfec08ee949411a5018c6171ea6dcdc7f372b8
+size 4952185256

model-00002-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dae79fd08d48f7385249cc05ebc563cd5c5d99d36d71c93e9fa91527622623d7
+size 4930227792

model-00003-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e58fe7684a0d6d2a35c047589e8e3915343dc01c3882fec032b49d67530e92c
+size 4927118464

model-00004-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d386a7d62fcb851beef1af136dd86efab0dc7dad95462d55c6d7f92c7cb643a
+size 4930227864

model-00005-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c7bd558c2e9d1ebe9f79d170d183bc50a6f4816632d429b8e77acd118854e49
+size 4927118464

model-00006-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f31f911dcfec4146007ffb5fdda36fbc9799d5b16b76c2e546e44572cb010bc7
+size 4930227864

model-00007-of-00007.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8cfd70249b9ea230b231ce078ac37d3b56265e3b655f3892a9d6035475ae829
+size 1437858576

model.safetensors.index.json

@@ -0,0 +1,491 @@
+{
+  "metadata": {
+    "total_size": 31034912768
+  },
+  "weight_map": {
+    "transformer.h.0.attn.c_attn.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.0.attn.c_attn.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.0.attn.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.0.attn.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.0.ln_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.0.ln_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.0.ln_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.0.ln_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.0.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.0.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.0.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.0.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.1.attn.c_attn.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.1.attn.c_attn.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.1.attn.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.1.attn.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.1.ln_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.1.ln_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.1.ln_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.1.ln_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.1.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.1.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.1.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.1.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.10.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.10.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.10.attn.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.10.attn.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.10.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.10.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.10.ln_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.10.ln_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.10.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.10.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.10.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.10.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.11.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.11.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.11.attn.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.11.attn.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.11.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.11.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.11.ln_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.11.ln_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.11.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.11.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.11.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.11.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.12.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.12.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.12.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.12.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.12.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.12.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.12.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.12.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.12.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.12.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.12.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.12.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.13.attn.c_attn.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.13.attn.c_attn.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.13.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.13.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.13.ln_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.13.ln_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.13.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.13.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.13.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.13.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.13.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.13.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.14.attn.c_attn.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.14.attn.c_attn.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.14.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.14.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.14.ln_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.14.ln_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.14.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.14.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.14.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.14.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.14.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.14.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.15.attn.c_attn.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.15.attn.c_attn.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.15.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.15.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.15.ln_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.15.ln_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.15.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.15.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.15.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.15.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.15.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.15.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.16.attn.c_attn.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.16.attn.c_attn.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.16.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.16.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.16.ln_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.16.ln_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.16.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.16.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.16.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.16.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.16.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.16.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.17.attn.c_attn.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.17.attn.c_attn.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.17.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.17.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.17.ln_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.17.ln_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.17.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.17.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.17.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.17.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.17.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.17.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.18.attn.c_attn.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.18.attn.c_attn.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.18.attn.c_proj.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.18.attn.c_proj.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.18.ln_1.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.18.ln_1.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.18.ln_2.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.18.ln_2.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.18.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
+    "transformer.h.18.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
+    "transformer.h.18.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.18.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.19.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.19.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.19.attn.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.19.attn.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.19.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.19.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.19.ln_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.19.ln_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.19.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.19.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.19.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.19.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.2.attn.c_attn.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.2.attn.c_attn.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.2.attn.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.2.attn.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.2.ln_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.2.ln_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.2.ln_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.2.ln_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.2.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.2.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.2.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.2.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.20.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.20.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.20.attn.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.20.attn.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.20.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.20.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.20.ln_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.20.ln_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.20.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.20.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.20.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.20.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.21.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.21.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.21.attn.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.21.attn.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.21.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.21.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.21.ln_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.21.ln_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.21.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.21.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.21.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.21.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.22.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.22.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.22.attn.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.22.attn.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.22.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.22.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.22.ln_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.22.ln_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.22.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.22.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.22.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.22.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.23.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.23.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.23.attn.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.23.attn.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.23.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.23.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.23.ln_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.23.ln_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.23.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.23.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.23.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.23.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.24.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.24.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.24.attn.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.24.attn.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.24.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.24.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.24.ln_2.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.24.ln_2.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.24.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.24.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.24.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.24.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.25.attn.c_attn.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.25.attn.c_attn.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.25.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.25.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.25.ln_1.bias": "model-00004-of-00007.safetensors",
+    "transformer.h.25.ln_1.weight": "model-00004-of-00007.safetensors",
+    "transformer.h.25.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.25.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.25.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.25.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.25.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.25.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.26.attn.c_attn.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.26.attn.c_attn.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.26.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.26.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.26.ln_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.26.ln_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.26.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.26.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.26.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.26.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.26.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.26.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.27.attn.c_attn.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.27.attn.c_attn.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.27.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.27.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.27.ln_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.27.ln_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.27.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.27.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.27.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.27.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.27.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.27.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.28.attn.c_attn.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.28.attn.c_attn.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.28.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.28.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.28.ln_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.28.ln_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.28.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.28.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.28.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.28.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.28.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.28.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.29.attn.c_attn.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.29.attn.c_attn.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.29.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.29.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.29.ln_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.29.ln_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.29.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.29.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.29.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.29.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.29.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.29.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.3.attn.c_attn.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.3.attn.c_attn.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.3.attn.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.3.attn.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.3.ln_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.3.ln_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.3.ln_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.3.ln_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.3.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.3.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.3.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.3.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.30.attn.c_attn.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.30.attn.c_attn.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.30.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.30.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.30.ln_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.30.ln_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.30.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.30.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.30.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.30.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.30.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.30.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.31.attn.c_attn.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.31.attn.c_attn.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.31.attn.c_proj.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.31.attn.c_proj.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.31.ln_1.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.31.ln_1.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.31.ln_2.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.31.ln_2.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.31.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
+    "transformer.h.31.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
+    "transformer.h.31.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.31.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.32.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.32.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.32.attn.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.32.attn.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.32.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.32.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.32.ln_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.32.ln_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.32.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.32.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.32.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.32.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.33.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.33.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.33.attn.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.33.attn.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.33.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.33.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.33.ln_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.33.ln_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.33.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.33.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.33.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.33.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.34.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.34.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.34.attn.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.34.attn.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.34.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.34.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.34.ln_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.34.ln_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.34.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.34.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.34.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.34.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.35.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.35.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.35.attn.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.35.attn.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.35.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.35.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.35.ln_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.35.ln_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.35.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.35.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.35.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.35.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.36.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.36.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.36.attn.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.36.attn.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.36.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.36.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.36.ln_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.36.ln_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.36.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.36.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.36.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.36.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.37.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.37.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.37.attn.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.37.attn.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.37.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.37.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.37.ln_2.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.37.ln_2.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.37.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.37.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.37.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.37.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.38.attn.c_attn.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.38.attn.c_attn.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.38.attn.c_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.38.attn.c_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.38.ln_1.bias": "model-00006-of-00007.safetensors",
+    "transformer.h.38.ln_1.weight": "model-00006-of-00007.safetensors",
+    "transformer.h.38.ln_2.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.38.ln_2.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.38.mlp.c_fc.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.38.mlp.c_fc.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.38.mlp.c_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.38.mlp.c_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.39.attn.c_attn.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.39.attn.c_attn.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.39.attn.c_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.39.attn.c_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.39.ln_1.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.39.ln_1.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.39.ln_2.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.39.ln_2.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.39.mlp.c_fc.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.39.mlp.c_fc.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.39.mlp.c_proj.bias": "model-00007-of-00007.safetensors",
+    "transformer.h.39.mlp.c_proj.weight": "model-00007-of-00007.safetensors",
+    "transformer.h.4.attn.c_attn.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.4.attn.c_attn.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.4.attn.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.4.attn.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.4.ln_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.4.ln_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.4.ln_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.4.ln_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.4.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.4.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.4.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.4.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.5.attn.c_attn.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.5.attn.c_attn.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.5.attn.c_proj.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.5.attn.c_proj.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.5.ln_1.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.5.ln_1.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.5.ln_2.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.5.ln_2.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.5.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
+    "transformer.h.5.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
+    "transformer.h.5.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.5.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.6.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.6.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.6.attn.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.6.attn.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.6.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.6.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.6.ln_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.6.ln_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.6.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.6.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.6.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.6.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.7.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.7.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.7.attn.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.7.attn.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.7.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.7.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.7.ln_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.7.ln_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.7.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.7.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.7.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.7.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.8.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.8.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.8.attn.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.8.attn.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.8.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.8.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.8.ln_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.8.ln_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.8.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.8.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.8.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.8.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.9.attn.c_attn.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.9.attn.c_attn.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.9.attn.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.9.attn.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.9.ln_1.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.9.ln_1.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.9.ln_2.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.9.ln_2.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.9.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.9.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
+    "transformer.h.9.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
+    "transformer.h.9.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
+    "transformer.ln_f.bias": "model-00007-of-00007.safetensors",
+    "transformer.ln_f.weight": "model-00007-of-00007.safetensors",
+    "transformer.wpe.weight": "model-00001-of-00007.safetensors",
+    "transformer.wte.weight": "model-00001-of-00007.safetensors"
+  }
+}