diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..52373fe24473b1aa44333d318f578ae6bf04b49b 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/LICENSE b/LICENSE new file mode 100644 index 0000000000000000000000000000000000000000..986b06fb2916b3518b82c7dcc0870723b91f2f92 --- /dev/null +++ b/LICENSE @@ -0,0 +1,21 @@ +MIT License + +Copyright (c) 2026 Z.AI Co., Ltd + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..719983da4516e02f056a479bf39719b744eb1f5c --- /dev/null +++ b/README.md @@ -0,0 +1,73 @@ +--- +library_name: mlx +license: mit +license_link: LICENSE +pipeline_tag: image-text-to-text +base_model: zai-org/GLM-5.3-Flash-BF16 +tags: +- mlx +--- + +# GLM-5.3-Flash-mlx-3bit-e3o8 + +[zai-org/GLM-5.3-Flash-BF16](https://huggingface.co/zai-org/GLM-5.3-Flash-BF16)(合計 320B・1 トークン 18B の MoE)の MLX 版。 +エキスパートだけを 3bit、ほかの線形層を 8bit にしたもの。手元の評価では、エキスパート 4bit 版と区別がつかなかった。 + +- サイズ: 134 GiB(143.9GB)。モデルサーバーのメモリは約 135GB(phys_footprint) +- 量子化(affine): + - ルーティングされるエキスパート(`switch_mlp.{gate,up,down}_proj`、42 層 × 3): **3bit g64** + - 注意機構・共有エキスパート・先頭 3 層の dense MLP・埋め込み・出力層など、ほかの線形層: **8bit g64** + - MoE のルーター(`mlp.gate`)・画像部分: 量子化しない(bf16) + - mHC・減衰の配列など: mlx-vlm の判定どおり fp32 + - MTP 層(層 45): mlx-vlm の変換と同じく含めない + +## 作り方 + +bf16(643GB)と出力を同時に置けないため、シャードを層ごとに落とし、mlx-vlm 0.7.11 の `convert` と同じ手順 +(sanitize → 型の変換 → `quantize_model`)を層ごとに当てて保存した。 +先頭 5 層だけに縮めたモデルで `mlx_vlm.convert` の出力と比べ、すべてのテンソル・config・ほかのファイルが一致することを確かめてある。 + +## 評価 + +`reasoning_effort: "low"`(後述)、温度 0.6・top_p 0.95・top_k 20、各ケース 5 回。mlx-vlm 0.7.11 のモデルサーバーを直接使い、MTP なし。 +probe は日本語・英語の短い問答 34 問、道具は呼び出しの場面 20 種を日本語と英語でそれぞれ。 + +| | e4o8(エキスパート 4bit・非公開) | **e3o8(このリポジトリ)** | e2o8(エキスパート 2bit g128・非公開) | +|---|---|---|---| +| 重みの大きさ | 181.9GB | 143.9GB | 96.3GB | +| モデルサーバーのメモリ | 約 170GB | 135GB | 91GB | +| probe 正答 | 100% | 100% | 90.4% | +| 日本語の道具 | 100% | 100% | 95% | +| 英語の道具 | 100% | 100% | 99% | +| 同じ句の繰り返しに陥る(probe) | 0% | 0% | 2.9% | +| 生成速度 | 29.6 tok/s | 30.3 tok/s | 31.8 tok/s | + +生成速度は M3 Ultra、温度 0・400 トークン × 3 問の平均。 + +- エキスパートを 2bit にすると壊れる(読みの誤り、事実の誤り、同じ句の繰り返し)。3bit が実用の下限。 +- ビットを下げても生成はほとんど速くならない。MoE では 1 トークンで読むエキスパートが一部だけなので、読む量の差が小さいため。 +- 長い会話・画像入力は試していない。 + +## 思考の切り替え + +チャットテンプレートに思考を切るスイッチは無く、`reasoning_effort`(low / high / max、既定は max)で深さを選ぶ。 +low にすると、ほぼ思考せずに答える(思考欄に短いメモが出ることはある)。 +mlx-vlm のサーバーでは、`reasoning_effort` をリクエストのトップレベルに置く(`chat_template_kwargs` の中に入れると効かない)。 + +``` +curl http://127.0.0.1:8080/v1/chat/completions -H 'Content-Type: application/json' -d '{ + "model": "ToPo-ToPo/GLM-5.3-Flash-mlx-3bit-e3o8", + "messages": [{"role": "user", "content": "こんにちは"}], + "reasoning_effort": "low" +}' +``` + +## 使い方 + +``` +pip install -U mlx-vlm +python -m mlx_vlm generate --model ToPo-ToPo/GLM-5.3-Flash-mlx-3bit-e3o8 \ + --prompt "日本の都道府県の数を教えてください" --max-tokens 512 +``` + +元のモデルのライセンス(MIT)に従う。 diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..14180451f166c1dab87081d8c3ee2a347f9d9d81 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,257 @@ +[gMASK] +{%- set effective_reasoning_effort = reasoning_effort if reasoning_effort is defined and reasoning_effort in ['low', 'high'] else 'max' -%} +{%- if effective_reasoning_effort is not none -%}<|system|>Reasoning Effort: {{ effective_reasoning_effort | capitalize }}{%- endif -%} +{%- set clear_thinking = clear_thinking if clear_thinking is defined else false -%} +{%- if tools -%} +{%- macro tool_to_json(tool) -%} + {%- set ns_tool = namespace(first=true) -%} + {{ '{' -}} + {%- for k, v in tool.items() -%} + {%- if k != 'defer_loading' and k != 'strict' -%} + {%- if not ns_tool.first -%}{{- ', ' -}}{%- endif -%} + {%- set ns_tool.first = false -%} + "{{ k }}": {{ v | tojson(ensure_ascii=False) }} + {%- endif -%} + {%- endfor -%} + {{- '}' -}} +{%- endmacro -%} +{%- macro tool_references_to_response(refs) -%} + {{- '\n' -}} + {%- for tr in refs -%} + {%- for tool in tools -%} + {%- if 'function' in tool -%} + {%- set tool = tool['function'] -%} + {%- endif -%} + {%- if tool.name == tr.name -%} + {{- tool_to_json(tool) + '\n' -}} + {%- endif -%} + {%- endfor -%} + {%- endfor -%} + {{- '' -}} +{%- endmacro -%} +<|system|> +# Tools + +You may call one or more functions to assist with the user query. + +You are provided with function signatures within XML tags: + +{% for tool in tools %} +{%- if 'function' in tool -%} + {%- set tool = tool['function'] -%} +{%- endif -%} +{% if tool.defer_loading is not defined or not tool.defer_loading %} +{{ tool_to_json(tool) }} +{% endif %} +{% endfor %} + + +For each function call, output the function name and arguments within the following XML format: +{function-name}{arg-key-1}{arg-value-1}{arg-key-2}{arg-value-2}...{%- endif -%} +{%- macro emit_image() -%}<|begin_of_image|><|image|><|end_of_image|>{%- endmacro -%} +{%- macro emit_video() -%}<|begin_of_video|><|video|><|end_of_video|>{%- endmacro -%} +{%- macro emit_audio() -%}<|begin_of_audio|><|end_of_audio|>{%- endmacro -%} +{%- macro visible_text(content) -%} + {%- if content is string -%} + {{- content -}} + {%- elif content is iterable and content is not mapping -%} + {%- for item in content -%} + {%- if item is mapping and item.type == 'text' -%} + {{- item.text -}} + {%- elif item is string -%} + {{- item -}} + {%- elif item is mapping and item.type in ['image', 'image_url'] -%} + {{- emit_image() -}} + {%- elif item is mapping and item.type in ['video', 'video_url'] -%} + {{- emit_video() -}} + {%- elif item is mapping and item.type in ['audio', 'audio_url', 'input_audio'] -%} + {{- emit_audio() -}} + {%- endif -%} + {%- endfor -%} + {%- elif content is not none -%} + {{- content -}} + {%- endif -%} +{%- endmacro -%} +{%- macro tool_response(text) -%} +{{- '' + text + '' -}} +{%- endmacro -%} +{%- macro render_tool_response(m) -%} +{%- if m.content is string -%} + {{- tool_response(m.content) -}} +{%- elif m.content and m.content is not mapping and m.content.0.type == "tool_reference" -%} + {{- tool_references_to_response(m.content) -}} +{%- elif is_list_of_outputs(m) -%} + {%- for tr in m.content -%} + {%- if tr.output is iterable and tr.output is not string and tr.output is not mapping and tr.output and tr.output.0.type == "tool_reference" -%} + {{- tool_references_to_response(tr.output) -}} + {%- else -%} + {{- tool_response(visible_text(tr.output)) -}} + {%- endif -%} + {%- endfor -%} +{%- else -%} + {{- tool_response(visible_text(m.content)) -}} +{%- endif -%} +{%- endmacro -%} +{%- macro id_of(obj) -%} + {%- if obj.tool_call_id -%} + {{- obj.tool_call_id -}} + {%- elif obj.id -%} + {{- obj.id -}} + {%- endif -%} +{%- endmacro -%} +{%- macro is_list_of_outputs(m) -%} + {%- if m.content and m.content.0.output is defined -%}1{%- endif -%} +{%- endmacro -%} +{%- macro has_dup_tool_result_id(lo, hi, target) -%} + {%- set ns_cnt = namespace(n=0) -%} + {%- for k in range(lo, hi + 1) -%} + {%- set m = messages[k] -%} + {%- if is_list_of_outputs(m) -%} + {%- for entry in m.content -%} + {%- if id_of(entry) == target -%} + {%- set ns_cnt.n = ns_cnt.n + 1 -%} + {%- endif -%} + {%- endfor -%} + {%- elif id_of(m) == target -%} + {%- set ns_cnt.n = ns_cnt.n + 1 -%} + {%- endif -%} + {%- if ns_cnt.n > 1 -%}{%- break -%}{%- endif -%} + {%- endfor -%} + {%- if ns_cnt.n > 1 -%}1{%- endif -%} +{%- endmacro -%} +{%- macro tc_id_exists(tcs, target) -%} + {%- set ns_f = namespace(found=false) -%} + {%- for tc in tcs -%} + {%- if id_of(tc) == target -%} + {%- set ns_f.found = true -%} + {%- break -%} + {%- endif -%} + {%- endfor -%} + {%- if ns_f.found -%}1{%- endif -%} +{%- endmacro -%} +{%- set ns = namespace(last_user_index=-1) -%} +{%- for m in messages %} + {%- if m.role == 'user' %} + {%- set ns.last_user_index = loop.index0 -%} + {%- endif %} +{%- endfor %} +{%- for m in messages -%} +{%- if m.role == 'user' -%}<|user|>{{ visible_text(m.content) }} +{%- elif m.role == 'assistant' -%} +<|assistant|> +{%- set content = visible_text(m.content) %} +{%- if m.reasoning_content is string %} + {%- set reasoning_content = m.reasoning_content %} +{%- elif '' in content %} + {%- set reasoning_content = content.split('')[0].split('')[-1] %} + {%- set content = content.split('')[-1] %} +{%- endif %} +{%- if (not clear_thinking or loop.index0 > ns.last_user_index) and reasoning_content is defined -%} +{{ '' + reasoning_content + ''}} +{%- else -%} +{{ '' }} +{%- endif -%} +{%- if content.strip() -%} +{{ content.strip() }} +{%- endif -%} +{% if m.tool_calls %} +{% for tc in m.tool_calls %} +{%- if tc.function %} + {%- set tc = tc.function %} +{%- endif %} +{{- '' ~ tc.name -}} +{% set _args = tc.arguments %}{% for k, v in _args.items() %}{{ k }}{{ v | tojson(ensure_ascii=False) if v is not string else v }}{% endfor %}{% endfor %} +{% endif %} +{%- elif m.role == 'tool' -%} +{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|observation|>' -}} + {%- set block_start = loop.index0 -%} + {%- set ns_blk = namespace(end=block_start) -%} + {%- for j in range(block_start, messages|length) -%} + {%- if messages[j].role == 'tool' -%} + {%- set ns_blk.end = j -%} + {%- else -%} + {%- break -%} + {%- endif -%} + {%- endfor -%} + {%- set ns_a = namespace(tool_calls=none) -%} + {%- if block_start > 0 and messages[block_start - 1].role == 'assistant' and messages[block_start - 1].tool_calls -%} + {%- set ns_a.tool_calls = messages[block_start - 1].tool_calls -%} + {%- endif -%} + {%- set ns_chk = namespace(can_sort=true, res=[], mk=[], ek=[], tcs=[]) -%} + {%- for k in range(block_start, ns_blk.end + 1) -%} + {%- set m = messages[k] -%} + {%- if is_list_of_outputs(m) -%} + {%- for entry in m.content -%} + {%- set ns_chk.res = ns_chk.res + [id_of(entry)] -%} + {%- set ns_chk.mk = ns_chk.mk + [k] -%} + {%- set ns_chk.ek = ns_chk.ek + [loop.index0] -%} + {%- endfor -%} + {%- else -%} + {%- set ns_chk.res = ns_chk.res + [id_of(m)] -%} + {%- set ns_chk.mk = ns_chk.mk + [k] -%} + {%- set ns_chk.ek = ns_chk.ek + [-1] -%} + {%- endif -%} + {%- endfor -%} + {%- if not ns_a.tool_calls -%} + {%- set ns_chk.can_sort = false -%} + {%- else -%} + {%- for tc in ns_a.tool_calls -%} + {%- set ns_chk.tcs = ns_chk.tcs + [id_of(tc)] -%} + {%- endfor -%} + {%- if '' in ns_chk.res or '' in ns_chk.tcs -%} + {%- set ns_chk.can_sort = false -%} + {%- elif ns_chk.res | unique(case_sensitive=true) | list | length != ns_chk.res | length -%} + {%- set ns_chk.can_sort = false -%} + {%- elif ns_chk.tcs | unique(case_sensitive=true) | list | length != ns_chk.tcs | length -%} + {%- set ns_chk.can_sort = false -%} + {%- else -%} + {%- for eid in ns_chk.res -%} + {%- if eid not in ns_chk.tcs -%} + {%- set ns_chk.can_sort = false -%} + {%- break -%} + {%- endif -%} + {%- endfor -%} + {%- endif -%} + {%- endif -%} + {%- if ns_chk.can_sort -%} + {%- set ns_ord = namespace(res=[], tcs=[], j=0, pairs=[]) -%} + {%- for eid in ns_chk.res -%} + {%- set ns_ord.res = ns_ord.res + [{'id': eid, 'p': loop.index0}] -%} + {%- endfor -%} + {%- for tc_id in ns_chk.tcs -%} + {%- set ns_ord.tcs = ns_ord.tcs + [{'id': tc_id, 'i': loop.index0}] -%} + {%- endfor -%} + {%- set ns_ord.res = ns_ord.res | sort(attribute='id', case_sensitive=true) -%} + {%- for tc in ns_ord.tcs | sort(attribute='id', case_sensitive=true) -%} + {%- if ns_ord.j < ns_ord.res | length and ns_ord.res[ns_ord.j].id == tc.id -%} + {%- set ns_ord.pairs = ns_ord.pairs + [{'i': tc.i, 'p': ns_ord.res[ns_ord.j].p}] -%} + {%- set ns_ord.j = ns_ord.j + 1 -%} + {%- endif -%} + {%- endfor -%} + {%- for pair in ns_ord.pairs | sort(attribute='i') -%} + {%- set m = messages[ns_chk.mk[pair.p]] -%} + {%- if ns_chk.ek[pair.p] >= 0 -%} + {%- set entry = m.content[ns_chk.ek[pair.p]] -%} + {%- if entry.output is iterable and entry.output is not string and entry.output is not mapping and entry.output and entry.output.0.type == "tool_reference" -%} + {{- tool_references_to_response(entry.output) -}} + {%- else -%} + {{- tool_response(visible_text(entry.output)) -}} + {%- endif -%} + {%- else -%} + {{- render_tool_response(m) -}} + {%- endif -%} + {%- endfor -%} + {%- else -%} + {%- for k in range(block_start, ns_blk.end + 1) -%} + {{- render_tool_response(messages[k]) -}} + {%- endfor -%} + {%- endif -%} +{% endif -%} +{%- elif m.role == 'system' -%} +<|system|>{{ visible_text(m.content) }} +{%- endif -%} +{%- endfor -%} +{%- if add_generation_prompt -%} + <|assistant|>{{- '' -}} +{%- endif -%} diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..67d169df1114c469db100bd4cdc51c53b4dfd78c --- /dev/null +++ b/config.json @@ -0,0 +1,4128 @@ +{ + "architectures": [ + "Glm5NextForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": [ + 154820, + 154827, + 154829 + ], + "generation_config": { + "_from_model_config": true, + "eos_token_id": [ + 154820, + 154827, + 154829 + ], + "pad_token_id": 154820, + "temperature": 1.0, + "top_p": 0.95, + "transformers_version": "5.16.0" + }, + "image_end_token_id": 154831, + "image_start_token_id": 154830, + "image_token_id": 154854, + "model_type": "glm5_next", + "quantization": { + "group_size": 64, + "bits": 3, + "mode": "affine", + "language_model.model.layers.0.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.mlp.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.mlp.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.mlp.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.mlp.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.10.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.10.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.10.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.embed_tokens": { + "group_size": 64, + "bits": 8 + }, + "language_model.lm_head": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.11.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.11.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.11.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.12.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.12.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.12.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.13.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.13.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.13.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.14.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.14.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.14.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.15.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.15.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.15.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.16.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.16.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.16.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.17.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.17.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.17.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.18.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.18.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.18.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.19.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.19.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.19.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.mlp.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.mlp.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.20.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.20.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.20.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.21.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.21.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.21.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.22.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.22.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.22.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.23.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.23.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.23.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.24.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.24.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.24.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.25.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.25.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.25.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.26.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.26.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.26.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.27.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.27.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.27.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.28.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.28.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.28.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.29.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.29.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.29.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.3.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.3.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.3.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.30.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.30.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.30.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.31.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.31.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.31.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.32.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.32.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.32.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.33.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.33.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.33.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.34.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.34.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.34.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.35.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.35.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.35.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.36.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.36.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.36.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.37.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.37.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.37.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.38.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.38.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.38.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.39.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.39.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.39.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.4.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.4.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.4.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.40.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.40.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.40.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.41.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.41.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.41.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.42.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.42.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.42.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.43.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.43.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.43.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.44.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.44.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.44.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.5.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.5.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.5.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.6.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.6.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.6.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.7.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.7.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.7.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.8.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.8.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.8.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.9.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.9.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.9.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + } + }, + "quantization_config": { + "group_size": 64, + "bits": 3, + "mode": "affine", + "language_model.model.layers.0.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.mlp.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.0.mlp.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.mlp.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.1.mlp.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.10.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.10.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.10.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.10.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.embed_tokens": { + "group_size": 64, + "bits": 8 + }, + "language_model.lm_head": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.11.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.11.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.11.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.11.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.12.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.12.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.12.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.12.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.13.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.13.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.13.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.13.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.14.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.14.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.14.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.14.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.15.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.15.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.15.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.15.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.16.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.16.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.16.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.16.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.17.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.17.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.17.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.17.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.18.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.18.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.18.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.18.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.19.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.19.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.19.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.19.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.mlp.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.2.mlp.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.20.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.20.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.20.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.20.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.21.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.21.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.21.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.21.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.22.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.22.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.22.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.22.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.23.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.23.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.23.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.23.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.24.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.24.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.24.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.24.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.25.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.25.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.25.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.25.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.26.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.26.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.26.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.26.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.27.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.27.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.27.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.27.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.28.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.28.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.28.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.28.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.29.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.29.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.29.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.29.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.3.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.3.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.3.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.3.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.30.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.30.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.30.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.30.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.31.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.31.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.31.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.31.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.32.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.32.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.32.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.32.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.33.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.33.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.33.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.33.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.34.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.34.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.34.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.34.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.35.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.35.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.35.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.35.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.36.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.36.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.36.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.36.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.37.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.37.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.37.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.37.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.38.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.38.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.38.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.38.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.39.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.39.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.39.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.39.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.4.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.4.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.4.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.4.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.40.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.40.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.40.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.40.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.41.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.41.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.41.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.41.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.42.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.42.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.42.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.42.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.43.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.43.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.43.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.43.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.44.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.44.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.44.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.44.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.5.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.5.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.5.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.5.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.6.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.6.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.6.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.6.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.qkv_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.q_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.embed_q": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.unembed_out": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.indexer.wq_b": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.indexer.wk": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.self_attn.indexer.weights_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.7.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.7.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.7.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.7.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.8.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.8.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.8.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.8.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.qkv_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.fbg_a_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.f_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.g_b_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.self_attn.o_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.mlp.switch_mlp.gate_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.9.mlp.switch_mlp.up_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.9.mlp.switch_mlp.down_proj": { + "group_size": 64, + "bits": 3 + }, + "language_model.model.layers.9.mlp.shared_experts.gate_up_proj": { + "group_size": 64, + "bits": 8 + }, + "language_model.model.layers.9.mlp.shared_experts.down_proj": { + "group_size": 64, + "bits": 8 + } + }, + "temperature": 1.0, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "dtype": "bfloat16", + "eos_token_id": [ + 154820, + 154827, + 154829 + ], + "first_k_dense_replace": 3, + "hc_eps": 1e-06, + "hc_mult": 4, + "hc_sinkhorn_iters": 20, + "head_dim": 0, + "hidden_act": "silu", + "hidden_size": 4096, + "index_head_dim": 128, + "index_kpool": 4, + "index_kpool_always_select_tail": true, + "index_kpool_compress": true, + "index_n_heads": 32, + "index_topk": 2048, + "index_share_for_mtp_iteration": true, + "indexer_rope_interleave": true, + "indexer_types": [ + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full", + "full" + ], + "initializer_range": 0.02, + "intermediate_size": 12288, + "kv_lora_rank": 512, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "deepseek_sparse_attention", + "linear_attention" + ], + "linear_attn_config": { + "num_heads": 64, + "gate_lower_bound": -5.0, + "head_dim": 128, + "short_conv_kernel_size": 4, + "kda_layers": [ + 0, + 1, + 2, + 4, + 5, + 6, + 8, + 9, + 10, + 12, + 13, + 14, + 16, + 17, + 18, + 20, + 21, + 22, + 24, + 25, + 26, + 28, + 29, + 30, + 32, + 33, + 34, + 36, + 37, + 38, + 40, + 41, + 42, + 44 + ], + "full_attn_layers": [ + 3, + 7, + 11, + 15, + 19, + 23, + 27, + 31, + 35, + 39, + 43 + ] + }, + "max_position_embeddings": 1048576, + "mhc": true, + "mla_use_nope": true, + "mlp_layer_types": [ + "dense", + "dense", + "dense", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse", + "sparse" + ], + "model_type": "glm5_next_text", + "moe_intermediate_size": 2048, + "moe_router_dtype": "float32", + "n_group": 1, + "n_routed_experts": 288, + "n_shared_experts": 1, + "norm_topk_prob": true, + "num_attention_heads": 64, + "num_experts_per_tok": 8, + "num_hidden_layers": 45, + "num_key_value_heads": 64, + "num_nextn_predict_layers": 1, + "output_router_logits": false, + "pad_token_id": 154820, + "q_lora_rank": 1536, + "qk_head_dim": 256, + "qk_nope_head_dim": 256, + "qk_rope_head_dim": 0, + "rms_norm_eps": 1e-05, + "routed_scaling_factor": 2.5, + "router_aux_loss_coef": 0.001, + "scoring_func": "sigmoid", + "swiglu_limit": 10.0, + "tie_word_embeddings": false, + "topk_group": 1, + "topk_method": "noaux_tc", + "use_cache": true, + "v_head_dim": 256, + "vocab_size": 154880 + }, + "tie_word_embeddings": false, + "top_p": 0.95, + "transformers_version": "5.16.0", + "video_end_token_id": 154833, + "video_start_token_id": 154832, + "video_token_id": 154855, + "vision_config": { + "attention_bias": true, + "attention_dropout": 0.0, + "depth": 24, + "hidden_act": "silu", + "hidden_size": 1024, + "image_size": 448, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "glm5_next_vision", + "num_heads": 16, + "out_hidden_size": 4096, + "patch_size": 14, + "projection_intermediate_size": 10240, + "rms_norm_eps": 1e-05, + "spatial_merge_size": 2, + "swiglu_limit": 10.0, + "temporal_patch_size": 2 + } +} \ No newline at end of file diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..637ee6afa5c6ebfba4eb24bd5b9df68550143597 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 154820, + 154827, + 154829 + ], + "pad_token_id": 154820, + "temperature": 1.0, + "top_p": 0.95, + "transformers_version": "5.16.0" +} diff --git a/model-00001-of-00046.safetensors b/model-00001-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b0850f7129b6b238cb7964cf8860c5175b7b51e6 --- /dev/null +++ b/model-00001-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3dd5a0d83e6f5c2dbc5bc54a48f58445df27324edf196c981ccaecb25b955e20 +size 308482767 diff --git a/model-00002-of-00046.safetensors b/model-00002-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f569bc1e4429018ef8c356b5be6e76d3f70f0fee --- /dev/null +++ b/model-00002-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d53998369d6c2dd3d3adaa250f40ac9f4445b092970ba2e31d584c5e8f4368e4 +size 308482809 diff --git a/model-00003-of-00046.safetensors b/model-00003-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d4771bdee302a82c883ca7de26190c0128fea64c --- /dev/null +++ b/model-00003-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c82369dca070719e29f263f36f4cdb6c07f9e0041e39de0560a17700dfb622f3 +size 3348045313 diff --git a/model-00004-of-00046.safetensors b/model-00004-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b078290d634560a068c4426727bbdca2c914667 --- /dev/null +++ b/model-00004-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a79a1e3e38df1569958d261f4cb84829dad99ce40ecd86cbcacb7141f84e154b +size 2475378128 diff --git a/model-00005-of-00046.safetensors b/model-00005-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6fa46be2d1d434e922e13638d10379c3d4a2c63e --- /dev/null +++ b/model-00005-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b9bd339b9ab41fc4c3ddb189f5fa5681697ccf158f82d36d726851e5cf285538 +size 3334805545 diff --git a/model-00006-of-00046.safetensors b/model-00006-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ed781d1ebedd8a38b45ac7f5e385fa6a68c6da7b --- /dev/null +++ b/model-00006-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4bd09b8229b32922bbb4cb40066f7a7ddd0c779bd3c0cef9b1b4484ba4ae762a +size 3348045355 diff --git a/model-00007-of-00046.safetensors b/model-00007-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f74d96028880acfc4ceb82094c791e561ed5441e --- /dev/null +++ b/model-00007-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd6dff45a997a15cc185a47e346949e84305f8f1011472a26f19ca221615af9c +size 3348045363 diff --git a/model-00008-of-00046.safetensors b/model-00008-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..418154ba243da1f333fa26852c4879b61e99011d --- /dev/null +++ b/model-00008-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc6c7147c303fa6f93d9438b7b173c4805ececc5f5adc2211092a574ecfa42bd +size 3348045325 diff --git a/model-00009-of-00046.safetensors b/model-00009-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..854a8274859d31842c9f98a39e6ca8ad8f5f543e --- /dev/null +++ b/model-00009-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac3a4dc3e9d28ecadd82e20d49b2d5d868a7759e8b6ce6e4f61b336bee0e3e24 +size 3334805591 diff --git a/model-00010-of-00046.safetensors b/model-00010-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5e7d599ab55b78e350731001161f9d0946e25ac2 --- /dev/null +++ b/model-00010-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a713ae0978e305f6727ed46324eb09ddc00e410e74da32f80de092f9cdfd6f33 +size 3348045329 diff --git a/model-00011-of-00046.safetensors b/model-00011-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..070bdf27ebebd812619f7c9c804138cfdf8c0595 --- /dev/null +++ b/model-00011-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90fe0f1ce58dde1a4616447cf5557322c0387a17c96081765cd744ff58463d3b +size 3348045347 diff --git a/model-00012-of-00046.safetensors b/model-00012-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..41645565044211f1272a784bb4c6125ffdbf9e4c --- /dev/null +++ b/model-00012-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:929178c6032eb0bb0d007e7f3481259eeaa4064d21c5db9e8702629414793fc8 +size 3348045317 diff --git a/model-00013-of-00046.safetensors b/model-00013-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..92d40b7ade27c9428c7c7b0ac8f252c1102b49b1 --- /dev/null +++ b/model-00013-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f502c76df7c0e8919035940bf727389221758d96d02a4d7f2d924db89feb1f5 +size 3334805583 diff --git a/model-00014-of-00046.safetensors b/model-00014-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..897aa14dc1cdd8c3c9f4c4919b7f68c36617770e --- /dev/null +++ b/model-00014-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be462e4c27e919fb3cc080780cfe706cebfb3153416d2065639f4a2bdce1f1dd +size 308482797 diff --git a/model-00015-of-00046.safetensors b/model-00015-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ad0f5ffd3e5ed5251fe754fffcf45a3cc278575a --- /dev/null +++ b/model-00015-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c2dd42967ec8663b58527f164ab4f411cc4991401c9a6983aac99c14246808b +size 3348045333 diff --git a/model-00016-of-00046.safetensors b/model-00016-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6c02d3858b655fb2361c888edb6491bbc1bc7335 --- /dev/null +++ b/model-00016-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8994c0a3e25338fdac06266ff89ed525a741e03568e9a4496a96737386cfa9c3 +size 3348045355 diff --git a/model-00017-of-00046.safetensors b/model-00017-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2b54d68a2897add4356359571b0daac215172dbb --- /dev/null +++ b/model-00017-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:705bd6376ec1e80b3a7948720ec2ec5d464d76d815b8d670d79bb1d96edeb646 +size 3348045329 diff --git a/model-00018-of-00046.safetensors b/model-00018-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..63306c170e05b63ba637c52e8b8e632512fb9d96 --- /dev/null +++ b/model-00018-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:80fbd709a79db11a7b37e240c241040c0b3095b7a7e292ce35d9693d7f9e8532 +size 3334805583 diff --git a/model-00019-of-00046.safetensors b/model-00019-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..413303cf4495e20f4c2abe1c1a7f29c84befe5f6 --- /dev/null +++ b/model-00019-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:423475c13d6c2dcc273be94537e2aaf7265fb62399b9e1d5f8cd06ab5d985eb2 +size 3348045339 diff --git a/model-00020-of-00046.safetensors b/model-00020-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..41da73f8564e5850ae5a0920c70d046b414ed405 --- /dev/null +++ b/model-00020-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cacafb90458def6d9194d404161dfaff9a1a53008867195fa43388e9f32d9975 +size 3348045339 diff --git a/model-00021-of-00046.safetensors b/model-00021-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c665fc19442bcf2d8b09ac5ae5c3b090ce0edf2f --- /dev/null +++ b/model-00021-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2067c9068fa274a77bc2a980a3792296e592aff70cbcc3559ec971befc61fe89 +size 3348045321 diff --git a/model-00023-of-00046.safetensors b/model-00023-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9cc4ee2a5862d85c4f6db93fef72cd2901984de5 --- /dev/null +++ b/model-00023-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b0d95614fa91dcaca4f2823eec3dafde16a1d6f6f8010b5d006032972cb606ae +size 3348045315 diff --git a/model-00024-of-00046.safetensors b/model-00024-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6cfd6325350b443d22237fd06b6de8ecbbe1822d --- /dev/null +++ b/model-00024-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ab359d9e9ac60afe2fc87115bdf943f8ac9ec72e55d88f97a96e103206e2a67 +size 3348045351 diff --git a/model-00025-of-00046.safetensors b/model-00025-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..99f1fa63775d553214d199c3a95d235649ff588f --- /dev/null +++ b/model-00025-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a9e0e32e49a2e264afb4f1669f88396ce4070dff552f7df999fd888d12a22f3 +size 3334805526 diff --git a/model-00026-of-00046.safetensors b/model-00026-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a5f79ca44aee63dc3f2869dc555690afddd36e8f --- /dev/null +++ b/model-00026-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d0a497e99c767b3074e3016bbf7a06fc59fbab23a2ae78dbcdd358388939f5b +size 3348045339 diff --git a/model-00029-of-00046.safetensors b/model-00029-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fbca33fa013f3f78985bbcf118c888da63524896 --- /dev/null +++ b/model-00029-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb0f364c0b71b899a9db46f55c55e051c9fa5ca252201780241d2f2196bf4f7b +size 3348045335 diff --git a/model-00031-of-00046.safetensors b/model-00031-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..54d69a333957c98003fabeb536692b82c97b170a --- /dev/null +++ b/model-00031-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6742116c24b16760379a78a9ad33be37eeee4d86d115dd42234ed01cad69ffa +size 3334805585 diff --git a/model-00032-of-00046.safetensors b/model-00032-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6ba78bfb4c187d15abb8a78d559fe5e8db1907a0 --- /dev/null +++ b/model-00032-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c8b01d7d65539c60e5fe17850427de448776d6e0db11f4f9e35ec2340dd69fc +size 3348045333 diff --git a/model-00033-of-00046.safetensors b/model-00033-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ef810a7a359bf67bc7e64cd60ad0e682ffad58e9 --- /dev/null +++ b/model-00033-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:04fad95fa84f152c87fce246417b024a2dc18447fe97d452357e1933ec60555e +size 3348045313 diff --git a/model-00034-of-00046.safetensors b/model-00034-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b77ef4a3fa46ea0c5d5f2a62497ad7d30c6218a3 --- /dev/null +++ b/model-00034-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf20f16ec4d14f3fedbcdce1848a915257ceeeb9584e895076901b22d62efe63 +size 3348045317 diff --git a/model-00036-of-00046.safetensors b/model-00036-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e96d8c8e23e99f26e75c9089ed23a27b1325617d --- /dev/null +++ b/model-00036-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b1438d25de24cb557e06229b8e270769287d9c8ee8eacc5393284ce8a07780c +size 3348045283 diff --git a/model-00037-of-00046.safetensors b/model-00037-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aa1e41b12b0f6b832f01176ae22476e9165ccbb9 --- /dev/null +++ b/model-00037-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3448bb900c781c37964f69914483d572993f82841eebe0e18b9bbcceebeb7d26 +size 3348045351 diff --git a/model-00038-of-00046.safetensors b/model-00038-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a4a1e370de8bdb53a8e20453d32f15d6e6b006bc --- /dev/null +++ b/model-00038-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:354d7216f340fbe45abee8071c69d10c420c258a2490b1929bbc1f559ca10ff8 +size 3348045343 diff --git a/model-00039-of-00046.safetensors b/model-00039-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ce2523aeaf98b6d51292dc15090c40495f1cf1b8 --- /dev/null +++ b/model-00039-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cc8c5a04c36e9dc293fad3dec761f51dffe211624728d96b8cc31567b1cc5f8 +size 3348045335 diff --git a/model-00040-of-00046.safetensors b/model-00040-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..53723334576ddbea7f4375699ddb6c6c20eb8a18 --- /dev/null +++ b/model-00040-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ac3323b7c04cee9a2d237fa577fd12e808730673ca637f979733b507b6a902d +size 3334805621 diff --git a/model-00041-of-00046.safetensors b/model-00041-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aeebc5072898221c0c4cb557e969fdf23717764a --- /dev/null +++ b/model-00041-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5483788f1467bf9db22818c0970c1016a1e9866f84a74de0fc74f6f88ece49d7 +size 3348045303 diff --git a/model-00042-of-00046.safetensors b/model-00042-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c5e21a4721a6b226e97a6891c670a5e3d6a14a55 --- /dev/null +++ b/model-00042-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aea0b69763601c36eb68a34b010d6615443168cee6b04dc77cdde02bec270bea +size 3348045271 diff --git a/model-00043-of-00046.safetensors b/model-00043-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5059a67ac704b9b51e68c9be3cd47b8430f55218 --- /dev/null +++ b/model-00043-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f1d1bce4373e138d5d1348d8aa6286ebdea0ebd481e09ba467e17b3eafbc6c7 +size 3348045271 diff --git a/model-00044-of-00046.safetensors b/model-00044-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..37b14ef639431a5e0e1465cc9a5c5e92e0a43d63 --- /dev/null +++ b/model-00044-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb7c2e317eb8a623f116a7777f2e59c2ae2ac40869da350b61a2559dc7ec9586 +size 3334805492 diff --git a/model-00045-of-00046.safetensors b/model-00045-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0d9e0d6a8e16326455f134d23f13526431d2b268 --- /dev/null +++ b/model-00045-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b64fec243834a80ea968360476b3d16ece2bb276937e61bbe21e8e6ee1e4673f +size 3348045321 diff --git a/model-00046-of-00046.safetensors b/model-00046-of-00046.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e572f3c47b81df67511cbb61950deb425b3b5618 --- /dev/null +++ b/model-00046-of-00046.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57d7a1890805eeab88746f3e12ff6492a2c786820092eea5ad1a25d6984987a8 +size 3348045291 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..46537e859db5aaf8a79c36ee908a65051a377b73 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,2429 @@ +{ + "metadata": { + "total_size": 143872778488 + }, + "weight_map": { + "language_model.lm_head.biases": "model-00004-of-00046.safetensors", + "language_model.lm_head.scales": "model-00004-of-00046.safetensors", + "language_model.lm_head.weight": "model-00004-of-00046.safetensors", + "language_model.model.embed_tokens.biases": "model-00004-of-00046.safetensors", + "language_model.model.embed_tokens.scales": "model-00004-of-00046.safetensors", + "language_model.model.embed_tokens.weight": "model-00004-of-00046.safetensors", + "language_model.model.layers.0.attn_hc.base": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.attn_hc.fn": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.attn_hc.scale": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.ffn_hc.base": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.ffn_hc.fn": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.ffn_hc.scale": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.mlp.down_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.mlp.down_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.mlp.gate_up_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.mlp.gate_up_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.A_log": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.dt_bias": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.f_b_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.f_b_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.f_b_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.fbg_a_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.fbg_a_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.fbg_a_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.g_b_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.g_b_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.g_b_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.o_norm.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.o_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.o_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.qkv_conv.conv.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.qkv_proj.biases": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.qkv_proj.scales": "model-00001-of-00046.safetensors", + "language_model.model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00046.safetensors", + "language_model.model.layers.1.attn_hc.base": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.attn_hc.fn": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.attn_hc.scale": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.ffn_hc.base": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.ffn_hc.fn": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.ffn_hc.scale": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.input_layernorm.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.mlp.down_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.mlp.down_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.mlp.down_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.mlp.gate_up_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.mlp.gate_up_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.mlp.gate_up_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.post_attention_layernorm.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.A_log": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.dt_bias": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.f_b_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.f_b_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.f_b_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.fbg_a_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.fbg_a_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.fbg_a_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.g_b_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.g_b_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.g_b_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.o_norm.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.o_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.o_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.o_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.qkv_conv.conv.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.qkv_proj.biases": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.qkv_proj.scales": "model-00002-of-00046.safetensors", + "language_model.model.layers.1.self_attn.qkv_proj.weight": "model-00002-of-00046.safetensors", + "language_model.model.layers.10.attn_hc.base": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.attn_hc.fn": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.attn_hc.scale": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.ffn_hc.base": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.ffn_hc.fn": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.ffn_hc.scale": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.input_layernorm.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.gate.e_score_correction_bias": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.gate.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.shared_experts.down_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.shared_experts.down_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.shared_experts.down_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.shared_experts.gate_up_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.shared_experts.gate_up_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.shared_experts.gate_up_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.down_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.down_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.down_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.gate_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.gate_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.gate_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.up_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.up_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.mlp.switch_mlp.up_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.post_attention_layernorm.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.A_log": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.dt_bias": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.f_b_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.f_b_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.f_b_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.fbg_a_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.fbg_a_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.fbg_a_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.g_b_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.g_b_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.g_b_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.o_norm.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.o_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.o_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.o_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.qkv_conv.conv.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.qkv_proj.biases": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.qkv_proj.scales": "model-00003-of-00046.safetensors", + "language_model.model.layers.10.self_attn.qkv_proj.weight": "model-00003-of-00046.safetensors", + "language_model.model.layers.11.attn_hc.base": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.attn_hc.fn": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.attn_hc.scale": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.ffn_hc.base": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.ffn_hc.fn": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.ffn_hc.scale": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.input_layernorm.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.gate.e_score_correction_bias": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.gate.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.shared_experts.down_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.shared_experts.down_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.shared_experts.down_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.shared_experts.gate_up_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.shared_experts.gate_up_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.shared_experts.gate_up_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.down_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.down_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.down_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.gate_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.gate_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.gate_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.up_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.up_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.mlp.switch_mlp.up_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.post_attention_layernorm.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.embed_q.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.embed_q.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.embed_q.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.index_kpool_compress_ape": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.index_kpool_compress_gate": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.k_norm.bias": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.k_norm.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.weights_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.weights_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.weights_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.wk.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.wk.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.wk.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.wq_b.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.wq_b.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.indexer.wq_b.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.kv_a_layernorm.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.o_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.o_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.o_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.q_a_layernorm.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.q_b_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.q_b_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.q_b_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.qkv_a_proj.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.qkv_a_proj.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.qkv_a_proj.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.unembed_out.biases": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.unembed_out.scales": "model-00005-of-00046.safetensors", + "language_model.model.layers.11.self_attn.unembed_out.weight": "model-00005-of-00046.safetensors", + "language_model.model.layers.12.attn_hc.base": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.attn_hc.fn": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.attn_hc.scale": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.ffn_hc.base": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.ffn_hc.fn": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.ffn_hc.scale": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.input_layernorm.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.gate.e_score_correction_bias": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.gate.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.shared_experts.down_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.shared_experts.down_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.shared_experts.down_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.shared_experts.gate_up_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.shared_experts.gate_up_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.shared_experts.gate_up_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.down_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.down_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.down_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.gate_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.gate_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.gate_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.up_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.up_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.mlp.switch_mlp.up_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.post_attention_layernorm.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.A_log": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.dt_bias": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.f_b_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.f_b_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.f_b_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.fbg_a_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.fbg_a_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.fbg_a_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.g_b_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.g_b_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.g_b_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.o_norm.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.o_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.o_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.o_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.qkv_conv.conv.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.qkv_proj.biases": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.qkv_proj.scales": "model-00006-of-00046.safetensors", + "language_model.model.layers.12.self_attn.qkv_proj.weight": "model-00006-of-00046.safetensors", + "language_model.model.layers.13.attn_hc.base": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.attn_hc.fn": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.attn_hc.scale": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.ffn_hc.base": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.ffn_hc.fn": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.ffn_hc.scale": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.input_layernorm.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.gate.e_score_correction_bias": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.gate.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.shared_experts.down_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.shared_experts.down_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.shared_experts.down_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.shared_experts.gate_up_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.shared_experts.gate_up_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.shared_experts.gate_up_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.down_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.down_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.down_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.gate_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.gate_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.gate_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.up_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.up_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.mlp.switch_mlp.up_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.post_attention_layernorm.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.A_log": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.dt_bias": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.f_b_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.f_b_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.f_b_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.fbg_a_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.fbg_a_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.fbg_a_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.g_b_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.g_b_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.g_b_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.o_norm.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.o_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.o_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.o_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.qkv_conv.conv.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.qkv_proj.biases": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.qkv_proj.scales": "model-00007-of-00046.safetensors", + "language_model.model.layers.13.self_attn.qkv_proj.weight": "model-00007-of-00046.safetensors", + "language_model.model.layers.14.attn_hc.base": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.attn_hc.fn": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.attn_hc.scale": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.ffn_hc.base": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.ffn_hc.fn": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.ffn_hc.scale": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.input_layernorm.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.gate.e_score_correction_bias": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.gate.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.shared_experts.down_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.shared_experts.down_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.shared_experts.down_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.shared_experts.gate_up_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.shared_experts.gate_up_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.shared_experts.gate_up_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.down_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.down_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.down_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.gate_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.gate_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.gate_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.up_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.up_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.mlp.switch_mlp.up_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.post_attention_layernorm.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.A_log": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.dt_bias": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.f_b_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.f_b_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.f_b_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.fbg_a_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.fbg_a_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.fbg_a_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.g_b_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.g_b_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.g_b_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.o_norm.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.o_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.o_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.o_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.qkv_conv.conv.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.qkv_proj.biases": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.qkv_proj.scales": "model-00008-of-00046.safetensors", + "language_model.model.layers.14.self_attn.qkv_proj.weight": "model-00008-of-00046.safetensors", + "language_model.model.layers.15.attn_hc.base": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.attn_hc.fn": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.attn_hc.scale": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.ffn_hc.base": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.ffn_hc.fn": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.ffn_hc.scale": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.input_layernorm.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.gate.e_score_correction_bias": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.gate.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.shared_experts.down_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.shared_experts.down_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.shared_experts.down_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.shared_experts.gate_up_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.shared_experts.gate_up_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.shared_experts.gate_up_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.down_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.down_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.down_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.gate_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.gate_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.gate_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.up_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.up_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.mlp.switch_mlp.up_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.post_attention_layernorm.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.embed_q.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.embed_q.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.embed_q.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.index_kpool_compress_ape": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.index_kpool_compress_gate": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.k_norm.bias": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.k_norm.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.weights_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.weights_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.weights_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.wk.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.wk.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.wk.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.wq_b.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.wq_b.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.indexer.wq_b.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.kv_a_layernorm.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.o_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.o_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.o_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.q_a_layernorm.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.q_b_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.q_b_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.q_b_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.qkv_a_proj.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.qkv_a_proj.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.qkv_a_proj.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.unembed_out.biases": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.unembed_out.scales": "model-00009-of-00046.safetensors", + "language_model.model.layers.15.self_attn.unembed_out.weight": "model-00009-of-00046.safetensors", + "language_model.model.layers.16.attn_hc.base": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.attn_hc.fn": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.attn_hc.scale": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.ffn_hc.base": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.ffn_hc.fn": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.ffn_hc.scale": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.input_layernorm.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.gate.e_score_correction_bias": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.gate.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.shared_experts.down_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.shared_experts.down_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.shared_experts.down_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.shared_experts.gate_up_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.shared_experts.gate_up_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.shared_experts.gate_up_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.down_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.down_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.down_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.gate_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.gate_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.gate_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.up_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.up_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.mlp.switch_mlp.up_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.post_attention_layernorm.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.A_log": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.dt_bias": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.f_b_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.f_b_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.f_b_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.fbg_a_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.fbg_a_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.fbg_a_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.g_b_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.g_b_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.g_b_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.o_norm.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.o_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.o_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.o_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.qkv_conv.conv.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.qkv_proj.biases": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.qkv_proj.scales": "model-00010-of-00046.safetensors", + "language_model.model.layers.16.self_attn.qkv_proj.weight": "model-00010-of-00046.safetensors", + "language_model.model.layers.17.attn_hc.base": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.attn_hc.fn": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.attn_hc.scale": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.ffn_hc.base": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.ffn_hc.fn": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.ffn_hc.scale": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.input_layernorm.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.gate.e_score_correction_bias": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.gate.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.shared_experts.down_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.shared_experts.down_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.shared_experts.down_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.shared_experts.gate_up_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.shared_experts.gate_up_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.shared_experts.gate_up_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.down_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.down_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.down_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.gate_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.gate_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.gate_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.up_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.up_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.mlp.switch_mlp.up_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.post_attention_layernorm.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.A_log": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.dt_bias": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.f_b_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.f_b_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.f_b_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.fbg_a_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.fbg_a_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.fbg_a_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.g_b_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.g_b_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.g_b_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.o_norm.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.o_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.o_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.o_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.qkv_conv.conv.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.qkv_proj.biases": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.qkv_proj.scales": "model-00011-of-00046.safetensors", + "language_model.model.layers.17.self_attn.qkv_proj.weight": "model-00011-of-00046.safetensors", + "language_model.model.layers.18.attn_hc.base": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.attn_hc.fn": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.attn_hc.scale": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.ffn_hc.base": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.ffn_hc.fn": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.ffn_hc.scale": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.input_layernorm.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.gate.e_score_correction_bias": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.gate.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.shared_experts.down_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.shared_experts.down_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.shared_experts.down_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.shared_experts.gate_up_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.shared_experts.gate_up_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.shared_experts.gate_up_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.down_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.down_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.down_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.gate_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.gate_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.gate_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.up_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.up_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.mlp.switch_mlp.up_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.post_attention_layernorm.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.A_log": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.dt_bias": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.f_b_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.f_b_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.f_b_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.fbg_a_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.fbg_a_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.fbg_a_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.g_b_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.g_b_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.g_b_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.o_norm.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.o_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.o_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.o_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.qkv_conv.conv.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.qkv_proj.biases": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.qkv_proj.scales": "model-00012-of-00046.safetensors", + "language_model.model.layers.18.self_attn.qkv_proj.weight": "model-00012-of-00046.safetensors", + "language_model.model.layers.19.attn_hc.base": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.attn_hc.fn": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.attn_hc.scale": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.ffn_hc.base": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.ffn_hc.fn": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.ffn_hc.scale": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.input_layernorm.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.gate.e_score_correction_bias": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.gate.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.shared_experts.down_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.shared_experts.down_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.shared_experts.down_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.shared_experts.gate_up_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.shared_experts.gate_up_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.shared_experts.gate_up_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.down_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.down_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.down_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.gate_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.gate_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.gate_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.up_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.up_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.mlp.switch_mlp.up_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.post_attention_layernorm.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.embed_q.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.embed_q.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.embed_q.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.index_kpool_compress_ape": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.index_kpool_compress_gate": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.k_norm.bias": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.k_norm.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.weights_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.weights_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.weights_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.wk.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.wk.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.wk.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.wq_b.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.wq_b.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.indexer.wq_b.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.kv_a_layernorm.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.o_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.o_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.o_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.q_a_layernorm.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.q_b_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.q_b_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.q_b_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.qkv_a_proj.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.qkv_a_proj.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.qkv_a_proj.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.unembed_out.biases": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.unembed_out.scales": "model-00013-of-00046.safetensors", + "language_model.model.layers.19.self_attn.unembed_out.weight": "model-00013-of-00046.safetensors", + "language_model.model.layers.2.attn_hc.base": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.attn_hc.fn": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.attn_hc.scale": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.ffn_hc.base": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.ffn_hc.fn": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.ffn_hc.scale": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.input_layernorm.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.mlp.down_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.mlp.down_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.mlp.down_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.mlp.gate_up_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.mlp.gate_up_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.mlp.gate_up_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.post_attention_layernorm.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.A_log": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.dt_bias": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.f_b_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.f_b_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.f_b_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.fbg_a_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.fbg_a_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.fbg_a_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.g_b_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.g_b_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.g_b_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.o_norm.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.o_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.o_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.o_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.qkv_conv.conv.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.qkv_proj.biases": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.qkv_proj.scales": "model-00014-of-00046.safetensors", + "language_model.model.layers.2.self_attn.qkv_proj.weight": "model-00014-of-00046.safetensors", + "language_model.model.layers.20.attn_hc.base": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.attn_hc.fn": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.attn_hc.scale": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.ffn_hc.base": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.ffn_hc.fn": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.ffn_hc.scale": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.input_layernorm.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.gate.e_score_correction_bias": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.gate.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.shared_experts.down_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.shared_experts.down_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.shared_experts.down_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.shared_experts.gate_up_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.shared_experts.gate_up_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.shared_experts.gate_up_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.down_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.down_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.down_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.gate_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.gate_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.gate_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.up_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.up_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.mlp.switch_mlp.up_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.post_attention_layernorm.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.A_log": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.dt_bias": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.f_b_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.f_b_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.f_b_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.fbg_a_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.fbg_a_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.fbg_a_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.g_b_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.g_b_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.g_b_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.o_norm.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.o_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.o_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.o_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.qkv_conv.conv.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.qkv_proj.biases": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.qkv_proj.scales": "model-00015-of-00046.safetensors", + "language_model.model.layers.20.self_attn.qkv_proj.weight": "model-00015-of-00046.safetensors", + "language_model.model.layers.21.attn_hc.base": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.attn_hc.fn": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.attn_hc.scale": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.ffn_hc.base": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.ffn_hc.fn": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.ffn_hc.scale": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.input_layernorm.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.gate.e_score_correction_bias": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.gate.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.shared_experts.down_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.shared_experts.down_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.shared_experts.down_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.shared_experts.gate_up_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.shared_experts.gate_up_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.shared_experts.gate_up_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.down_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.down_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.down_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.gate_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.gate_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.gate_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.up_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.up_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.mlp.switch_mlp.up_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.post_attention_layernorm.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.A_log": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.dt_bias": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.f_b_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.f_b_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.f_b_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.fbg_a_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.fbg_a_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.fbg_a_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.g_b_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.g_b_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.g_b_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.o_norm.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.o_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.o_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.o_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.qkv_conv.conv.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.qkv_proj.biases": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.qkv_proj.scales": "model-00016-of-00046.safetensors", + "language_model.model.layers.21.self_attn.qkv_proj.weight": "model-00016-of-00046.safetensors", + "language_model.model.layers.22.attn_hc.base": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.attn_hc.fn": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.attn_hc.scale": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.ffn_hc.base": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.ffn_hc.fn": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.ffn_hc.scale": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.input_layernorm.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.gate.e_score_correction_bias": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.gate.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.shared_experts.down_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.shared_experts.down_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.shared_experts.down_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.shared_experts.gate_up_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.shared_experts.gate_up_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.shared_experts.gate_up_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.down_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.down_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.down_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.gate_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.gate_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.gate_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.up_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.up_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.mlp.switch_mlp.up_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.post_attention_layernorm.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.A_log": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.dt_bias": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.f_b_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.f_b_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.f_b_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.fbg_a_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.fbg_a_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.fbg_a_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.g_b_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.g_b_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.g_b_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.o_norm.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.o_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.o_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.o_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.qkv_conv.conv.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.qkv_proj.biases": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.qkv_proj.scales": "model-00017-of-00046.safetensors", + "language_model.model.layers.22.self_attn.qkv_proj.weight": "model-00017-of-00046.safetensors", + "language_model.model.layers.23.attn_hc.base": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.attn_hc.fn": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.attn_hc.scale": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.ffn_hc.base": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.ffn_hc.fn": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.ffn_hc.scale": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.input_layernorm.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.gate.e_score_correction_bias": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.gate.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.shared_experts.down_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.shared_experts.down_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.shared_experts.down_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.shared_experts.gate_up_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.shared_experts.gate_up_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.shared_experts.gate_up_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.down_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.down_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.down_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.gate_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.gate_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.gate_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.up_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.up_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.mlp.switch_mlp.up_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.post_attention_layernorm.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.embed_q.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.embed_q.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.embed_q.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.index_kpool_compress_ape": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.index_kpool_compress_gate": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.k_norm.bias": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.k_norm.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.weights_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.weights_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.weights_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.wk.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.wk.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.wk.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.wq_b.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.wq_b.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.indexer.wq_b.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.kv_a_layernorm.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.o_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.o_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.o_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.q_a_layernorm.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.q_b_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.q_b_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.q_b_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.qkv_a_proj.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.qkv_a_proj.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.qkv_a_proj.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.unembed_out.biases": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.unembed_out.scales": "model-00018-of-00046.safetensors", + "language_model.model.layers.23.self_attn.unembed_out.weight": "model-00018-of-00046.safetensors", + "language_model.model.layers.24.attn_hc.base": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.attn_hc.fn": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.attn_hc.scale": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.ffn_hc.base": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.ffn_hc.fn": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.ffn_hc.scale": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.input_layernorm.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.gate.e_score_correction_bias": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.gate.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.shared_experts.down_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.shared_experts.down_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.shared_experts.down_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.shared_experts.gate_up_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.shared_experts.gate_up_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.shared_experts.gate_up_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.down_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.down_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.down_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.gate_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.gate_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.gate_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.up_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.up_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.mlp.switch_mlp.up_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.post_attention_layernorm.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.A_log": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.dt_bias": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.f_b_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.f_b_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.f_b_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.fbg_a_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.fbg_a_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.fbg_a_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.g_b_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.g_b_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.g_b_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.o_norm.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.o_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.o_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.o_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.qkv_conv.conv.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.qkv_proj.biases": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.qkv_proj.scales": "model-00019-of-00046.safetensors", + "language_model.model.layers.24.self_attn.qkv_proj.weight": "model-00019-of-00046.safetensors", + "language_model.model.layers.25.attn_hc.base": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.attn_hc.fn": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.attn_hc.scale": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.ffn_hc.base": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.ffn_hc.fn": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.ffn_hc.scale": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.input_layernorm.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.gate.e_score_correction_bias": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.gate.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.shared_experts.down_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.shared_experts.down_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.shared_experts.down_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.shared_experts.gate_up_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.shared_experts.gate_up_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.shared_experts.gate_up_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.down_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.down_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.down_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.gate_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.gate_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.gate_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.up_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.up_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.mlp.switch_mlp.up_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.post_attention_layernorm.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.A_log": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.dt_bias": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.f_b_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.f_b_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.f_b_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.fbg_a_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.fbg_a_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.fbg_a_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.g_b_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.g_b_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.g_b_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.o_norm.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.o_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.o_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.o_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.qkv_conv.conv.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.qkv_proj.biases": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.qkv_proj.scales": "model-00020-of-00046.safetensors", + "language_model.model.layers.25.self_attn.qkv_proj.weight": "model-00020-of-00046.safetensors", + "language_model.model.layers.26.attn_hc.base": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.attn_hc.fn": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.attn_hc.scale": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.ffn_hc.base": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.ffn_hc.fn": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.ffn_hc.scale": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.input_layernorm.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.gate.e_score_correction_bias": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.gate.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.shared_experts.down_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.shared_experts.down_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.shared_experts.down_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.shared_experts.gate_up_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.shared_experts.gate_up_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.shared_experts.gate_up_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.down_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.down_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.down_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.gate_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.gate_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.gate_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.up_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.up_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.mlp.switch_mlp.up_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.post_attention_layernorm.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.A_log": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.dt_bias": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.f_b_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.f_b_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.f_b_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.fbg_a_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.fbg_a_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.fbg_a_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.g_b_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.g_b_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.g_b_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.o_norm.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.o_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.o_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.o_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.qkv_conv.conv.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.qkv_proj.biases": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.qkv_proj.scales": "model-00021-of-00046.safetensors", + "language_model.model.layers.26.self_attn.qkv_proj.weight": "model-00021-of-00046.safetensors", + "language_model.model.layers.27.attn_hc.base": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.attn_hc.fn": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.attn_hc.scale": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.ffn_hc.base": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.ffn_hc.fn": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.ffn_hc.scale": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.input_layernorm.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.gate.e_score_correction_bias": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.gate.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.shared_experts.down_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.shared_experts.down_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.shared_experts.down_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.shared_experts.gate_up_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.shared_experts.gate_up_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.shared_experts.gate_up_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.down_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.down_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.down_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.gate_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.gate_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.gate_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.up_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.up_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.mlp.switch_mlp.up_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.post_attention_layernorm.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.embed_q.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.embed_q.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.embed_q.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.index_kpool_compress_ape": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.index_kpool_compress_gate": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.k_norm.bias": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.k_norm.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.weights_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.weights_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.weights_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.wk.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.wk.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.wk.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.wq_b.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.wq_b.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.indexer.wq_b.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.kv_a_layernorm.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.o_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.o_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.o_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.q_a_layernorm.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.q_b_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.q_b_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.q_b_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.qkv_a_proj.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.qkv_a_proj.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.qkv_a_proj.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.unembed_out.biases": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.unembed_out.scales": "model-00022-of-00046.safetensors", + "language_model.model.layers.27.self_attn.unembed_out.weight": "model-00022-of-00046.safetensors", + "language_model.model.layers.28.attn_hc.base": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.attn_hc.fn": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.attn_hc.scale": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.ffn_hc.base": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.ffn_hc.fn": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.ffn_hc.scale": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.input_layernorm.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.gate.e_score_correction_bias": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.gate.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.shared_experts.down_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.shared_experts.down_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.shared_experts.down_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.shared_experts.gate_up_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.shared_experts.gate_up_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.shared_experts.gate_up_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.down_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.down_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.down_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.gate_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.gate_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.gate_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.up_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.up_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.mlp.switch_mlp.up_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.post_attention_layernorm.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.A_log": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.dt_bias": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.f_b_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.f_b_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.f_b_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.fbg_a_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.fbg_a_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.fbg_a_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.g_b_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.g_b_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.g_b_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.o_norm.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.o_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.o_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.o_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.qkv_conv.conv.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.qkv_proj.biases": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.qkv_proj.scales": "model-00023-of-00046.safetensors", + "language_model.model.layers.28.self_attn.qkv_proj.weight": "model-00023-of-00046.safetensors", + "language_model.model.layers.29.attn_hc.base": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.attn_hc.fn": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.attn_hc.scale": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.ffn_hc.base": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.ffn_hc.fn": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.ffn_hc.scale": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.input_layernorm.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.gate.e_score_correction_bias": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.gate.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.shared_experts.down_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.shared_experts.down_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.shared_experts.down_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.shared_experts.gate_up_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.shared_experts.gate_up_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.shared_experts.gate_up_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.down_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.down_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.down_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.gate_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.gate_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.gate_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.up_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.up_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.mlp.switch_mlp.up_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.post_attention_layernorm.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.A_log": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.dt_bias": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.f_b_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.f_b_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.f_b_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.fbg_a_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.fbg_a_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.fbg_a_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.g_b_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.g_b_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.g_b_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.o_norm.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.o_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.o_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.o_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.qkv_conv.conv.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.qkv_proj.biases": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.qkv_proj.scales": "model-00024-of-00046.safetensors", + "language_model.model.layers.29.self_attn.qkv_proj.weight": "model-00024-of-00046.safetensors", + "language_model.model.layers.3.attn_hc.base": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.attn_hc.fn": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.attn_hc.scale": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.ffn_hc.base": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.ffn_hc.fn": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.ffn_hc.scale": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.input_layernorm.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.gate.e_score_correction_bias": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.gate.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.shared_experts.down_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.shared_experts.down_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.shared_experts.down_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.shared_experts.gate_up_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.shared_experts.gate_up_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.shared_experts.gate_up_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.down_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.down_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.down_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.gate_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.gate_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.gate_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.up_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.up_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.mlp.switch_mlp.up_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.post_attention_layernorm.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.embed_q.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.embed_q.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.embed_q.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.index_kpool_compress_ape": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.index_kpool_compress_gate": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.k_norm.bias": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.k_norm.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.weights_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.weights_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.weights_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.wk.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.wk.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.wk.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.wq_b.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.wq_b.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.indexer.wq_b.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.kv_a_layernorm.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.o_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.o_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.o_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.q_a_layernorm.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.q_b_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.q_b_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.q_b_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.qkv_a_proj.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.qkv_a_proj.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.qkv_a_proj.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.unembed_out.biases": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.unembed_out.scales": "model-00025-of-00046.safetensors", + "language_model.model.layers.3.self_attn.unembed_out.weight": "model-00025-of-00046.safetensors", + "language_model.model.layers.30.attn_hc.base": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.attn_hc.fn": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.attn_hc.scale": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.ffn_hc.base": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.ffn_hc.fn": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.ffn_hc.scale": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.input_layernorm.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.gate.e_score_correction_bias": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.gate.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.shared_experts.down_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.shared_experts.down_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.shared_experts.down_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.shared_experts.gate_up_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.shared_experts.gate_up_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.shared_experts.gate_up_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.down_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.down_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.down_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.gate_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.gate_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.gate_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.up_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.up_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.mlp.switch_mlp.up_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.post_attention_layernorm.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.A_log": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.dt_bias": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.f_b_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.f_b_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.f_b_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.fbg_a_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.fbg_a_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.fbg_a_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.g_b_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.g_b_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.g_b_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.o_norm.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.o_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.o_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.o_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.qkv_conv.conv.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.qkv_proj.biases": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.qkv_proj.scales": "model-00026-of-00046.safetensors", + "language_model.model.layers.30.self_attn.qkv_proj.weight": "model-00026-of-00046.safetensors", + "language_model.model.layers.31.attn_hc.base": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.attn_hc.fn": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.attn_hc.scale": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.ffn_hc.base": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.ffn_hc.fn": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.ffn_hc.scale": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.input_layernorm.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.gate.e_score_correction_bias": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.gate.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.shared_experts.down_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.shared_experts.down_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.shared_experts.down_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.shared_experts.gate_up_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.shared_experts.gate_up_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.shared_experts.gate_up_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.down_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.down_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.down_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.gate_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.gate_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.gate_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.up_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.up_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.mlp.switch_mlp.up_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.post_attention_layernorm.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.embed_q.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.embed_q.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.embed_q.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.index_kpool_compress_ape": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.index_kpool_compress_gate": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.k_norm.bias": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.k_norm.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.weights_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.weights_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.weights_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.wk.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.wk.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.wk.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.wq_b.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.wq_b.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.indexer.wq_b.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.kv_a_layernorm.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.o_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.o_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.o_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.q_a_layernorm.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.q_b_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.q_b_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.q_b_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.qkv_a_proj.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.qkv_a_proj.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.qkv_a_proj.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.unembed_out.biases": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.unembed_out.scales": "model-00027-of-00046.safetensors", + "language_model.model.layers.31.self_attn.unembed_out.weight": "model-00027-of-00046.safetensors", + "language_model.model.layers.32.attn_hc.base": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.attn_hc.fn": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.attn_hc.scale": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.ffn_hc.base": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.ffn_hc.fn": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.ffn_hc.scale": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.input_layernorm.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.gate.e_score_correction_bias": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.gate.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.shared_experts.down_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.shared_experts.down_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.shared_experts.down_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.shared_experts.gate_up_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.shared_experts.gate_up_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.shared_experts.gate_up_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.down_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.down_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.down_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.gate_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.gate_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.gate_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.up_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.up_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.mlp.switch_mlp.up_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.post_attention_layernorm.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.A_log": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.dt_bias": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.f_b_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.f_b_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.f_b_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.fbg_a_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.fbg_a_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.fbg_a_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.g_b_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.g_b_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.g_b_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.o_norm.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.o_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.o_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.o_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.qkv_conv.conv.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.qkv_proj.biases": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.qkv_proj.scales": "model-00028-of-00046.safetensors", + "language_model.model.layers.32.self_attn.qkv_proj.weight": "model-00028-of-00046.safetensors", + "language_model.model.layers.33.attn_hc.base": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.attn_hc.fn": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.attn_hc.scale": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.ffn_hc.base": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.ffn_hc.fn": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.ffn_hc.scale": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.input_layernorm.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.gate.e_score_correction_bias": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.gate.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.shared_experts.down_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.shared_experts.down_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.shared_experts.down_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.shared_experts.gate_up_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.shared_experts.gate_up_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.shared_experts.gate_up_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.down_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.down_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.down_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.gate_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.gate_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.gate_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.up_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.up_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.mlp.switch_mlp.up_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.post_attention_layernorm.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.A_log": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.dt_bias": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.f_b_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.f_b_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.f_b_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.fbg_a_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.fbg_a_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.fbg_a_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.g_b_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.g_b_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.g_b_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.o_norm.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.o_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.o_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.o_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.qkv_conv.conv.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.qkv_proj.biases": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.qkv_proj.scales": "model-00029-of-00046.safetensors", + "language_model.model.layers.33.self_attn.qkv_proj.weight": "model-00029-of-00046.safetensors", + "language_model.model.layers.34.attn_hc.base": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.attn_hc.fn": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.attn_hc.scale": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.ffn_hc.base": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.ffn_hc.fn": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.ffn_hc.scale": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.input_layernorm.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.gate.e_score_correction_bias": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.gate.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.shared_experts.down_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.shared_experts.down_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.shared_experts.down_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.shared_experts.gate_up_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.shared_experts.gate_up_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.shared_experts.gate_up_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.down_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.down_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.down_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.gate_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.gate_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.gate_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.up_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.up_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.mlp.switch_mlp.up_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.post_attention_layernorm.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.A_log": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.dt_bias": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.f_b_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.f_b_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.f_b_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.fbg_a_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.fbg_a_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.fbg_a_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.g_b_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.g_b_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.g_b_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.o_norm.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.o_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.o_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.o_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.qkv_conv.conv.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.qkv_proj.biases": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.qkv_proj.scales": "model-00030-of-00046.safetensors", + "language_model.model.layers.34.self_attn.qkv_proj.weight": "model-00030-of-00046.safetensors", + "language_model.model.layers.35.attn_hc.base": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.attn_hc.fn": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.attn_hc.scale": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.ffn_hc.base": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.ffn_hc.fn": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.ffn_hc.scale": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.input_layernorm.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.gate.e_score_correction_bias": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.gate.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.shared_experts.down_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.shared_experts.down_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.shared_experts.down_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.shared_experts.gate_up_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.shared_experts.gate_up_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.shared_experts.gate_up_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.down_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.down_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.down_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.gate_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.gate_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.gate_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.up_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.up_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.mlp.switch_mlp.up_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.post_attention_layernorm.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.embed_q.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.embed_q.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.embed_q.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.index_kpool_compress_ape": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.index_kpool_compress_gate": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.k_norm.bias": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.k_norm.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.weights_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.weights_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.weights_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.wk.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.wk.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.wk.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.wq_b.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.wq_b.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.indexer.wq_b.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.kv_a_layernorm.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.o_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.o_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.o_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.q_a_layernorm.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.q_b_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.q_b_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.q_b_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.qkv_a_proj.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.qkv_a_proj.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.qkv_a_proj.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.unembed_out.biases": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.unembed_out.scales": "model-00031-of-00046.safetensors", + "language_model.model.layers.35.self_attn.unembed_out.weight": "model-00031-of-00046.safetensors", + "language_model.model.layers.36.attn_hc.base": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.attn_hc.fn": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.attn_hc.scale": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.ffn_hc.base": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.ffn_hc.fn": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.ffn_hc.scale": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.input_layernorm.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.gate.e_score_correction_bias": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.gate.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.shared_experts.down_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.shared_experts.down_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.shared_experts.down_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.shared_experts.gate_up_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.shared_experts.gate_up_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.shared_experts.gate_up_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.down_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.down_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.down_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.gate_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.gate_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.gate_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.up_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.up_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.mlp.switch_mlp.up_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.post_attention_layernorm.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.A_log": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.dt_bias": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.f_b_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.f_b_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.f_b_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.fbg_a_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.fbg_a_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.fbg_a_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.g_b_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.g_b_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.g_b_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.o_norm.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.o_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.o_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.o_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.qkv_conv.conv.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.qkv_proj.biases": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.qkv_proj.scales": "model-00032-of-00046.safetensors", + "language_model.model.layers.36.self_attn.qkv_proj.weight": "model-00032-of-00046.safetensors", + "language_model.model.layers.37.attn_hc.base": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.attn_hc.fn": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.attn_hc.scale": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.ffn_hc.base": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.ffn_hc.fn": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.ffn_hc.scale": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.input_layernorm.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.gate.e_score_correction_bias": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.gate.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.shared_experts.down_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.shared_experts.down_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.shared_experts.down_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.shared_experts.gate_up_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.shared_experts.gate_up_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.shared_experts.gate_up_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.down_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.down_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.down_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.gate_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.gate_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.gate_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.up_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.up_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.mlp.switch_mlp.up_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.post_attention_layernorm.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.A_log": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.dt_bias": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.f_b_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.f_b_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.f_b_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.fbg_a_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.fbg_a_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.fbg_a_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.g_b_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.g_b_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.g_b_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.o_norm.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.o_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.o_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.o_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.qkv_conv.conv.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.qkv_proj.biases": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.qkv_proj.scales": "model-00033-of-00046.safetensors", + "language_model.model.layers.37.self_attn.qkv_proj.weight": "model-00033-of-00046.safetensors", + "language_model.model.layers.38.attn_hc.base": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.attn_hc.fn": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.attn_hc.scale": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.ffn_hc.base": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.ffn_hc.fn": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.ffn_hc.scale": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.input_layernorm.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.gate.e_score_correction_bias": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.gate.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.shared_experts.down_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.shared_experts.down_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.shared_experts.down_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.shared_experts.gate_up_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.shared_experts.gate_up_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.shared_experts.gate_up_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.down_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.down_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.down_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.gate_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.gate_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.gate_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.up_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.up_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.mlp.switch_mlp.up_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.post_attention_layernorm.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.A_log": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.dt_bias": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.f_b_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.f_b_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.f_b_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.fbg_a_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.fbg_a_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.fbg_a_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.g_b_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.g_b_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.g_b_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.o_norm.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.o_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.o_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.o_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.qkv_conv.conv.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.qkv_proj.biases": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.qkv_proj.scales": "model-00034-of-00046.safetensors", + "language_model.model.layers.38.self_attn.qkv_proj.weight": "model-00034-of-00046.safetensors", + "language_model.model.layers.39.attn_hc.base": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.attn_hc.fn": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.attn_hc.scale": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.ffn_hc.base": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.ffn_hc.fn": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.ffn_hc.scale": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.input_layernorm.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.gate.e_score_correction_bias": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.gate.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.shared_experts.down_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.shared_experts.down_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.shared_experts.down_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.shared_experts.gate_up_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.shared_experts.gate_up_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.shared_experts.gate_up_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.down_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.down_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.down_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.gate_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.gate_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.gate_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.up_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.up_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.mlp.switch_mlp.up_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.post_attention_layernorm.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.embed_q.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.embed_q.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.embed_q.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.index_kpool_compress_ape": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.index_kpool_compress_gate": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.k_norm.bias": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.k_norm.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.weights_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.weights_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.weights_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.wk.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.wk.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.wk.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.wq_b.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.wq_b.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.indexer.wq_b.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.kv_a_layernorm.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.o_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.o_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.o_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.q_a_layernorm.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.q_b_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.q_b_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.q_b_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.qkv_a_proj.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.qkv_a_proj.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.qkv_a_proj.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.unembed_out.biases": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.unembed_out.scales": "model-00035-of-00046.safetensors", + "language_model.model.layers.39.self_attn.unembed_out.weight": "model-00035-of-00046.safetensors", + "language_model.model.layers.4.attn_hc.base": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.attn_hc.fn": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.attn_hc.scale": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.ffn_hc.base": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.ffn_hc.fn": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.ffn_hc.scale": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.input_layernorm.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.gate.e_score_correction_bias": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.gate.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.shared_experts.down_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.shared_experts.down_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.shared_experts.down_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.shared_experts.gate_up_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.shared_experts.gate_up_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.shared_experts.gate_up_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.down_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.down_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.down_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.gate_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.gate_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.gate_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.up_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.up_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.mlp.switch_mlp.up_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.post_attention_layernorm.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.A_log": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.dt_bias": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.f_b_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.f_b_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.f_b_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.fbg_a_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.fbg_a_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.fbg_a_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.g_b_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.g_b_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.g_b_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.o_norm.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.o_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.o_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.o_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.qkv_conv.conv.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.qkv_proj.biases": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.qkv_proj.scales": "model-00036-of-00046.safetensors", + "language_model.model.layers.4.self_attn.qkv_proj.weight": "model-00036-of-00046.safetensors", + "language_model.model.layers.40.attn_hc.base": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.attn_hc.fn": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.attn_hc.scale": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.ffn_hc.base": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.ffn_hc.fn": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.ffn_hc.scale": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.input_layernorm.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.gate.e_score_correction_bias": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.gate.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.shared_experts.down_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.shared_experts.down_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.shared_experts.down_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.shared_experts.gate_up_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.shared_experts.gate_up_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.shared_experts.gate_up_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.down_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.down_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.down_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.gate_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.gate_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.gate_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.up_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.up_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.mlp.switch_mlp.up_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.post_attention_layernorm.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.A_log": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.dt_bias": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.f_b_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.f_b_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.f_b_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.fbg_a_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.fbg_a_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.fbg_a_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.g_b_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.g_b_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.g_b_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.o_norm.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.o_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.o_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.o_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.qkv_conv.conv.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.qkv_proj.biases": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.qkv_proj.scales": "model-00037-of-00046.safetensors", + "language_model.model.layers.40.self_attn.qkv_proj.weight": "model-00037-of-00046.safetensors", + "language_model.model.layers.41.attn_hc.base": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.attn_hc.fn": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.attn_hc.scale": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.ffn_hc.base": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.ffn_hc.fn": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.ffn_hc.scale": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.input_layernorm.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.gate.e_score_correction_bias": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.gate.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.shared_experts.down_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.shared_experts.down_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.shared_experts.down_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.shared_experts.gate_up_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.shared_experts.gate_up_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.shared_experts.gate_up_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.down_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.down_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.down_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.gate_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.gate_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.gate_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.up_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.up_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.mlp.switch_mlp.up_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.post_attention_layernorm.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.A_log": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.dt_bias": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.f_b_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.f_b_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.f_b_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.fbg_a_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.fbg_a_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.fbg_a_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.g_b_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.g_b_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.g_b_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.o_norm.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.o_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.o_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.o_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.qkv_conv.conv.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.qkv_proj.biases": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.qkv_proj.scales": "model-00038-of-00046.safetensors", + "language_model.model.layers.41.self_attn.qkv_proj.weight": "model-00038-of-00046.safetensors", + "language_model.model.layers.42.attn_hc.base": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.attn_hc.fn": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.attn_hc.scale": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.ffn_hc.base": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.ffn_hc.fn": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.ffn_hc.scale": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.input_layernorm.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.gate.e_score_correction_bias": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.gate.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.shared_experts.down_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.shared_experts.down_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.shared_experts.down_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.shared_experts.gate_up_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.shared_experts.gate_up_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.shared_experts.gate_up_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.down_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.down_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.down_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.gate_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.gate_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.gate_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.up_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.up_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.mlp.switch_mlp.up_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.post_attention_layernorm.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.A_log": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.dt_bias": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.f_b_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.f_b_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.f_b_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.fbg_a_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.fbg_a_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.fbg_a_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.g_b_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.g_b_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.g_b_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.o_norm.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.o_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.o_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.o_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.qkv_conv.conv.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.qkv_proj.biases": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.qkv_proj.scales": "model-00039-of-00046.safetensors", + "language_model.model.layers.42.self_attn.qkv_proj.weight": "model-00039-of-00046.safetensors", + "language_model.model.layers.43.attn_hc.base": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.attn_hc.fn": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.attn_hc.scale": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.ffn_hc.base": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.ffn_hc.fn": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.ffn_hc.scale": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.input_layernorm.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.gate.e_score_correction_bias": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.gate.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.shared_experts.down_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.shared_experts.down_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.shared_experts.down_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.shared_experts.gate_up_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.shared_experts.gate_up_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.shared_experts.gate_up_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.down_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.down_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.down_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.gate_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.gate_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.gate_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.up_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.up_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.mlp.switch_mlp.up_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.post_attention_layernorm.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.embed_q.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.embed_q.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.embed_q.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.index_kpool_compress_ape": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.index_kpool_compress_gate": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.k_norm.bias": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.k_norm.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.weights_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.weights_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.weights_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.wk.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.wk.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.wk.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.wq_b.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.wq_b.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.indexer.wq_b.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.kv_a_layernorm.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.o_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.o_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.o_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.q_a_layernorm.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.q_b_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.q_b_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.q_b_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.qkv_a_proj.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.qkv_a_proj.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.qkv_a_proj.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.unembed_out.biases": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.unembed_out.scales": "model-00040-of-00046.safetensors", + "language_model.model.layers.43.self_attn.unembed_out.weight": "model-00040-of-00046.safetensors", + "language_model.model.layers.44.attn_hc.base": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.attn_hc.fn": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.attn_hc.scale": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.ffn_hc.base": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.ffn_hc.fn": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.ffn_hc.scale": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.input_layernorm.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.gate.e_score_correction_bias": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.gate.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.shared_experts.down_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.shared_experts.down_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.shared_experts.down_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.shared_experts.gate_up_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.shared_experts.gate_up_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.shared_experts.gate_up_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.down_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.down_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.down_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.gate_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.gate_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.gate_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.up_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.up_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.mlp.switch_mlp.up_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.post_attention_layernorm.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.A_log": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.dt_bias": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.f_b_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.f_b_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.f_b_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.fbg_a_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.fbg_a_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.fbg_a_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.g_b_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.g_b_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.g_b_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.o_norm.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.o_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.o_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.o_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.qkv_conv.conv.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.qkv_proj.biases": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.qkv_proj.scales": "model-00041-of-00046.safetensors", + "language_model.model.layers.44.self_attn.qkv_proj.weight": "model-00041-of-00046.safetensors", + "language_model.model.layers.5.attn_hc.base": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.attn_hc.fn": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.attn_hc.scale": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.ffn_hc.base": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.ffn_hc.fn": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.ffn_hc.scale": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.input_layernorm.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.gate.e_score_correction_bias": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.gate.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.shared_experts.down_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.shared_experts.down_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.shared_experts.down_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.shared_experts.gate_up_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.shared_experts.gate_up_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.shared_experts.gate_up_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.down_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.down_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.down_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.gate_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.gate_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.gate_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.up_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.up_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.mlp.switch_mlp.up_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.post_attention_layernorm.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.A_log": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.dt_bias": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.f_b_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.f_b_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.f_b_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.fbg_a_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.fbg_a_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.fbg_a_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.g_b_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.g_b_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.g_b_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.o_norm.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.o_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.o_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.o_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.qkv_conv.conv.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.qkv_proj.biases": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.qkv_proj.scales": "model-00042-of-00046.safetensors", + "language_model.model.layers.5.self_attn.qkv_proj.weight": "model-00042-of-00046.safetensors", + "language_model.model.layers.6.attn_hc.base": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.attn_hc.fn": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.attn_hc.scale": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.ffn_hc.base": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.ffn_hc.fn": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.ffn_hc.scale": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.input_layernorm.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.gate.e_score_correction_bias": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.gate.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.shared_experts.down_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.shared_experts.down_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.shared_experts.down_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.shared_experts.gate_up_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.shared_experts.gate_up_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.shared_experts.gate_up_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.down_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.down_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.down_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.gate_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.gate_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.gate_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.up_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.up_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.mlp.switch_mlp.up_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.post_attention_layernorm.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.A_log": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.dt_bias": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.f_b_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.f_b_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.f_b_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.fbg_a_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.fbg_a_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.fbg_a_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.g_b_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.g_b_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.g_b_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.o_norm.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.o_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.o_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.o_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.qkv_conv.conv.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.qkv_proj.biases": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.qkv_proj.scales": "model-00043-of-00046.safetensors", + "language_model.model.layers.6.self_attn.qkv_proj.weight": "model-00043-of-00046.safetensors", + "language_model.model.layers.7.attn_hc.base": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.attn_hc.fn": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.attn_hc.scale": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.ffn_hc.base": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.ffn_hc.fn": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.ffn_hc.scale": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.input_layernorm.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.gate.e_score_correction_bias": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.gate.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.shared_experts.down_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.shared_experts.down_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.shared_experts.down_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.shared_experts.gate_up_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.shared_experts.gate_up_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.shared_experts.gate_up_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.down_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.down_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.down_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.gate_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.gate_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.gate_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.up_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.up_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.mlp.switch_mlp.up_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.post_attention_layernorm.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.embed_q.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.embed_q.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.embed_q.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.index_kpool_compress_ape": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.index_kpool_compress_gate": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.k_norm.bias": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.k_norm.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.weights_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.weights_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.weights_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.wk.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.wk.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.wk.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.wq_b.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.wq_b.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.indexer.wq_b.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.kv_a_layernorm.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.o_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.o_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.o_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.q_a_layernorm.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.q_b_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.q_b_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.q_b_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.qkv_a_proj.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.qkv_a_proj.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.qkv_a_proj.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.unembed_out.biases": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.unembed_out.scales": "model-00044-of-00046.safetensors", + "language_model.model.layers.7.self_attn.unembed_out.weight": "model-00044-of-00046.safetensors", + "language_model.model.layers.8.attn_hc.base": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.attn_hc.fn": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.attn_hc.scale": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.ffn_hc.base": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.ffn_hc.fn": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.ffn_hc.scale": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.input_layernorm.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.gate.e_score_correction_bias": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.gate.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.shared_experts.down_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.shared_experts.down_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.shared_experts.down_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.shared_experts.gate_up_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.shared_experts.gate_up_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.shared_experts.gate_up_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.down_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.down_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.down_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.gate_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.gate_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.gate_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.up_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.up_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.mlp.switch_mlp.up_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.post_attention_layernorm.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.A_log": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.dt_bias": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.f_b_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.f_b_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.f_b_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.fbg_a_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.fbg_a_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.fbg_a_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.g_b_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.g_b_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.g_b_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.o_norm.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.o_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.o_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.o_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.qkv_conv.conv.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.qkv_proj.biases": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.qkv_proj.scales": "model-00045-of-00046.safetensors", + "language_model.model.layers.8.self_attn.qkv_proj.weight": "model-00045-of-00046.safetensors", + "language_model.model.layers.9.attn_hc.base": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.attn_hc.fn": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.attn_hc.scale": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.ffn_hc.base": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.ffn_hc.fn": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.ffn_hc.scale": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.input_layernorm.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.gate.e_score_correction_bias": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.gate.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.shared_experts.down_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.shared_experts.down_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.shared_experts.down_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.shared_experts.gate_up_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.shared_experts.gate_up_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.shared_experts.gate_up_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.down_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.down_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.down_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.gate_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.gate_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.gate_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.up_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.up_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.mlp.switch_mlp.up_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.post_attention_layernorm.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.A_log": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.dt_bias": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.f_b_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.f_b_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.f_b_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.fbg_a_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.fbg_a_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.fbg_a_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.g_b_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.g_b_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.g_b_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.o_norm.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.o_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.o_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.o_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.qkv_conv.conv.weight": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.qkv_proj.biases": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.qkv_proj.scales": "model-00046-of-00046.safetensors", + "language_model.model.layers.9.self_attn.qkv_proj.weight": "model-00046-of-00046.safetensors", + "language_model.model.norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.0.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.1.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.10.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.11.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.12.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.13.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.14.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.15.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.16.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.17.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.18.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.19.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.2.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.20.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.21.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.22.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.23.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.3.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.4.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.5.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.6.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.7.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.8.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.attn.k_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.attn.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.attn.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.attn.q_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.attn.qkv.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.attn.qkv.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.mlp.down_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.mlp.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.mlp.gate_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.mlp.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.mlp.up_proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.mlp.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.norm1.weight": "model-00004-of-00046.safetensors", + "vision_tower.blocks.9.norm2.weight": "model-00004-of-00046.safetensors", + "vision_tower.downsample.bias": "model-00004-of-00046.safetensors", + "vision_tower.downsample.weight": "model-00004-of-00046.safetensors", + "vision_tower.merger.down_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.merger.gate_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.merger.post_projection_norm.bias": "model-00004-of-00046.safetensors", + "vision_tower.merger.post_projection_norm.weight": "model-00004-of-00046.safetensors", + "vision_tower.merger.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.merger.up_proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.patch_embed.proj.bias": "model-00004-of-00046.safetensors", + "vision_tower.patch_embed.proj.weight": "model-00004-of-00046.safetensors", + "vision_tower.post_layernorm.weight": "model-00004-of-00046.safetensors" + } +} \ No newline at end of file diff --git a/processor_config.json b/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..44a1ef8bb6cf0356c92bad5372f12300b7b8b8af --- /dev/null +++ b/processor_config.json @@ -0,0 +1,69 @@ +{ + "image_processor": { + "default_to_square": false, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.48145466, + 0.4578275, + 0.40821073 + ], + "image_processor_type": "Glm5NextImageProcessor", + "image_std": [ + 0.26862954, + 0.26130258, + 0.27577711 + ], + "max_image_tokens": 8000, + "merge_size": 2, + "min_image_tokens": 16, + "patch_expand_factor": 1, + "patch_size": 14, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 1 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Glm5NextProcessor", + "video_processor": { + "default_to_square": false, + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.48145466, + 0.4578275, + 0.40821073 + ], + "image_processor_type": "Glm5NextVideoProcessor", + "image_std": [ + 0.26862954, + 0.26130258, + 0.27577711 + ], + "max_duration": 0, + "max_frames": 2048, + "max_image_size": { + "longest_edge": 47040000 + }, + "max_image_tokens": 240000, + "merge_size": 2, + "min_image_tokens": 16, + "num_frames": 16, + "patch_expand_factor": 1, + "patch_size": 14, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 1 + }, + "temporal_patch_size": 2 + } +} diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..aba40197a4cdb5607f4ab7a05fb0a4ee8054fd6d --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19e773648cb4e65de8660ea6365e10acca112d42a854923df93db4a6f333a82d +size 20217442 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..78aedebbbe1f8b96f366ee9064d7d3b524d12b0b --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,35 @@ +{ + "backend": "tokenizers", + "clean_up_tokenization_spaces": false, + "do_lower_case": false, + "eos_token": "<|endoftext|>", + "extra_special_tokens": [ + "<|endoftext|>", + "[MASK]", + "[gMASK]", + "[sMASK]", + "", + "", + "<|system|>", + "<|user|>", + "<|assistant|>", + "<|observation|>", + "<|begin_of_image|>", + "<|end_of_image|>", + "<|begin_of_video|>", + "<|end_of_video|>", + "<|begin_of_audio|>", + "<|end_of_audio|>", + "<|begin_of_transcription|>", + "<|end_of_transcription|>" + ], + "is_local": true, + "local_files_only": false, + "model_max_length": 1048576, + "model_specific_special_tokens": {}, + "pad_token": "<|endoftext|>", + "padding_side": "left", + "processor_class": "Glm5NextProcessor", + "remove_space": false, + "tokenizer_class": "TokenizersBackend" +}