Training in progress, step 800

.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

chat_template.jinja

@@ -0,0 +1,331 @@
+{#-
+  In addition to the normal inputs of `messages` and `tools`, this template also accepts the
+  following kwargs:
+  - "builtin_tools": A list, can contain "browser" and/or "python".
+  - "model_identity": A string that optionally describes the model identity.
+  - "reasoning_effort": A string that describes the reasoning effort, defaults to "medium".
+ #}
+
+{#- Tool Definition Rendering ============================================== #}
+{%- macro render_typescript_type(param_spec, required_params, is_nullable=false) -%}
+    {%- if param_spec.type == "array" -%}
+        {%- if param_spec['items'] -%}
+            {%- if param_spec['items']['type'] == "string" -%}
+                {{- "string[]" }}
+            {%- elif param_spec['items']['type'] == "number" -%}
+                {{- "number[]" }}
+            {%- elif param_spec['items']['type'] == "integer" -%}
+                {{- "number[]" }}
+            {%- elif param_spec['items']['type'] == "boolean" -%}
+                {{- "boolean[]" }}
+            {%- else -%}
+                {%- set inner_type = render_typescript_type(param_spec['items'], required_params) -%}
+                {%- if inner_type == "object | object" or inner_type|length > 50 -%}
+                    {{- "any[]" }}
+                {%- else -%}
+                    {{- inner_type + "[]" }}
+                {%- endif -%}
+            {%- endif -%}
+            {%- if param_spec.nullable -%}
+                {{- " | null" }}
+            {%- endif -%}
+        {%- else -%}
+            {{- "any[]" }}
+            {%- if param_spec.nullable -%}
+                {{- " | null" }}
+            {%- endif -%}
+        {%- endif -%}
+    {%- elif param_spec.type is defined and param_spec.type is iterable and param_spec.type is not string and param_spec.type is not mapping and param_spec.type[0] is defined -%}
+        {#- Handle array of types like ["object", "object"] from Union[dict, list] #}
+        {%- if param_spec.type | length > 1 -%}
+            {{- param_spec.type | join(" | ") }}
+        {%- else -%}
+            {{- param_spec.type[0] }}
+        {%- endif -%}
+    {%- elif param_spec.oneOf -%}
+        {#- Handle oneOf schemas - check for complex unions and fallback to any #}
+        {%- set has_object_variants = false -%}
+        {%- for variant in param_spec.oneOf -%}
+            {%- if variant.type == "object" -%}
+                {%- set has_object_variants = true -%}
+            {%- endif -%}
+        {%- endfor -%}
+        {%- if has_object_variants and param_spec.oneOf|length > 1 -%}
+            {{- "any" }}
+        {%- else -%}
+            {%- for variant in param_spec.oneOf -%}
+                {{- render_typescript_type(variant, required_params) -}}
+                {%- if variant.description %}
+                    {{- "// " + variant.description }}
+                {%- endif -%}
+                {%- if variant.default is defined %}
+                    {{ "// default: " + variant.default|tojson }}
+                {%- endif -%}
+                {%- if not loop.last %}
+                    {{- " | " }}
+                {% endif -%}
+            {%- endfor -%}
+        {%- endif -%}
+    {%- elif param_spec.type == "string" -%}
+        {%- if param_spec.enum -%}
+            {{- '"' + param_spec.enum|join('" | "') + '"' -}}
+        {%- else -%}
+            {{- "string" }}
+            {%- if param_spec.nullable %}
+                {{- " | null" }}
+            {%- endif -%}
+        {%- endif -%}
+    {%- elif param_spec.type == "number" -%}
+        {{- "number" }}
+    {%- elif param_spec.type == "integer" -%}
+        {{- "number" }}
+    {%- elif param_spec.type == "boolean" -%}
+        {{- "boolean" }}
+
+    {%- elif param_spec.type == "object" -%}
+        {%- if param_spec.properties -%}
+            {{- "{\n" }}
+            {%- for prop_name, prop_spec in param_spec.properties.items() -%}
+                {{- prop_name -}}
+                {%- if prop_name not in (param_spec.required or []) -%}
+                    {{- "?" }}
+                {%- endif -%}
+                {{- ": " }}
+                {{ render_typescript_type(prop_spec, param_spec.required or []) }}
+                {%- if not loop.last -%}
+                    {{-", " }}
+                {%- endif -%}
+            {%- endfor -%}
+            {{- "}" }}
+        {%- else -%}
+            {{- "object" }}
+        {%- endif -%}
+    {%- else -%}
+        {{- "any" }}
+    {%- endif -%}
+{%- endmacro -%}
+
+{%- macro render_tool_namespace(namespace_name, tools) -%}
+    {{- "## " + namespace_name + "\n\n" }}
+    {{- "namespace " + namespace_name + " {\n\n" }}
+    {%- for tool in tools %}
+        {%- set tool = tool.function %}
+        {{- "// " + tool.description + "\n" }}
+        {{- "type "+ tool.name + " = " }}
+        {%- if tool.parameters and tool.parameters.properties %}
+            {{- "(_: {\n" }}
+            {%- for param_name, param_spec in tool.parameters.properties.items() %}
+                {%- if param_spec.description %}
+                    {{- "// " + param_spec.description + "\n" }}
+                {%- endif %}
+                {{- param_name }}
+                {%- if param_name not in (tool.parameters.required or []) -%}
+                    {{- "?" }}
+                {%- endif -%}
+                {{- ": " }}
+                {{- render_typescript_type(param_spec, tool.parameters.required or []) }}
+                {%- if param_spec.default is defined -%}
+                    {%- if param_spec.enum %}
+                        {{- ", // default: " + param_spec.default }}
+                    {%- elif param_spec.oneOf %}
+                        {{- "// default: " + param_spec.default }}
+                    {%- else %}
+                        {{- ", // default: " + param_spec.default|tojson }}
+                    {%- endif -%}
+                {%- endif -%}
+                {%- if not loop.last %}
+                    {{- ",\n" }}
+                {%- else %}
+                    {{- ",\n" }}
+                {%- endif -%}
+            {%- endfor %}
+            {{- "}) => any;\n\n" }}
+        {%- else -%}
+            {{- "() => any;\n\n" }}
+        {%- endif -%}
+    {%- endfor %}
+    {{- "} // namespace " + namespace_name }}
+{%- endmacro -%}
+
+{%- macro render_builtin_tools(browser_tool, python_tool) -%}
+    {%- if browser_tool %}
+        {{- "## browser\n\n" }}
+        {{- "// Tool for browsing.\n" }}
+        {{- "// The `cursor` appears in brackets before each browsing display: `[{cursor}]`.\n" }}
+        {{- "// Cite information from the tool using the following format:\n" }}
+        {{- "// `【{cursor}†L{line_start}(-L{line_end})?】`, for example: `【6†L9-L11】` or `【8†L3】`.\n" }}
+        {{- "// Do not quote more than 10 words directly from the tool output.\n" }}
+        {{- "// sources=web (default: web)\n" }}
+        {{- "namespace browser {\n\n" }}
+        {{- "// Searches for information related to `query` and displays `topn` results.\n" }}
+        {{- "type search = (_: {\n" }}
+        {{- "query: string,\n" }}
+        {{- "topn?: number, // default: 10\n" }}
+        {{- "source?: string,\n" }}
+        {{- "}) => any;\n\n" }}
+        {{- "// Opens the link `id` from the page indicated by `cursor` starting at line number `loc`, showing `num_lines` lines.\n" }}
+        {{- "// Valid link ids are displayed with the formatting: `【{id}†.*】`.\n" }}
+        {{- "// If `cursor` is not provided, the most recent page is implied.\n" }}
+        {{- "// If `id` is a string, it is treated as a fully qualified URL associated with `source`.\n" }}
+        {{- "// If `loc` is not provided, the viewport will be positioned at the beginning of the document or centered on the most relevant passage, if available.\n" }}
+        {{- "// Use this function without `id` to scroll to a new location of an opened page.\n" }}
+        {{- "type open = (_: {\n" }}
+        {{- "id?: number | string, // default: -1\n" }}
+        {{- "cursor?: number, // default: -1\n" }}
+        {{- "loc?: number, // default: -1\n" }}
+        {{- "num_lines?: number, // default: -1\n" }}
+        {{- "view_source?: boolean, // default: false\n" }}
+        {{- "source?: string,\n" }}
+        {{- "}) => any;\n\n" }}
+        {{- "// Finds exact matches of `pattern` in the current page, or the page given by `cursor`.\n" }}
+        {{- "type find = (_: {\n" }}
+        {{- "pattern: string,\n" }}
+        {{- "cursor?: number, // default: -1\n" }}
+        {{- "}) => any;\n\n" }}
+        {{- "} // namespace browser\n\n" }}
+    {%- endif -%}
+
+    {%- if python_tool %}
+        {{- "## python\n\n" }}
+        {{- "Use this tool to execute Python code in your chain of thought. The code will not be shown to the user. This tool should be used for internal reasoning, but not for code that is intended to be visible to the user (e.g. when creating plots, tables, or files).\n\n" }}
+        {{- "When you send a message containing Python code to python, it will be executed in a stateful Jupyter notebook environment. python will respond with the output of the execution or time out after 120.0 seconds. The drive at '/mnt/data' can be used to save and persist user files. Internet access for this session is UNKNOWN. Depends on the cluster.\n\n" }}
+    {%- endif -%}
+{%- endmacro -%}
+
+{#- System Message Construction ============================================ #}
+{%- macro build_system_message() -%}
+    {%- if model_identity is not defined %}
+        {%- set model_identity = "You are ChatGPT, a large language model trained by OpenAI." %}
+    {%- endif %}
+    {{- model_identity + "\n" }}
+    {{- "Knowledge cutoff: 2024-06\n" }}
+    {{- "Current date: " + strftime_now("%Y-%m-%d") + "\n\n" }}
+    {%- if reasoning_effort is not defined %}
+        {%- set reasoning_effort = "medium" %}
+    {%- endif %}
+    {{- "Reasoning: " + reasoning_effort + "\n\n" }}
+    {%- if builtin_tools %}
+        {{- "# Tools\n\n" }}
+        {%- set available_builtin_tools = namespace(browser=false, python=false) %}
+        {%- for tool in builtin_tools %}
+            {%- if tool == "browser" %}
+                {%- set available_builtin_tools.browser = true %}
+            {%- elif tool == "python" %}
+                {%- set available_builtin_tools.python = true %}
+            {%- endif %}
+        {%- endfor %}
+        {{- render_builtin_tools(available_builtin_tools.browser, available_builtin_tools.python) }}
+    {%- endif -%}
+    {{- "# Valid channels: analysis, commentary, final. Channel must be included for every message." }}
+    {%- if tools -%}
+        {{- "\nCalls to these tools must go to the commentary channel: 'functions'." }}
+    {%- endif -%}
+{%- endmacro -%}
+
+{#- Main Template Logic ================================================= #}
+{#- Set defaults #}
+
+{#- Render system message #}
+{{- "<|start|>system<|message|>" }}
+{{- build_system_message() }}
+{{- "<|end|>" }}
+
+{#- Extract developer message #}
+{%- if messages[0].role == "developer" or messages[0].role == "system" %}
+    {%- set developer_message = messages[0].content %}
+    {%- set loop_messages = messages[1:] %}
+{%- else %}
+    {%- set developer_message = "" %}
+    {%- set loop_messages = messages %}
+{%- endif %}
+
+{#- Render developer message #}
+{%- if developer_message or tools %}
+    {{- "<|start|>developer<|message|>" }}
+    {%- if developer_message %}
+        {{- "# Instructions\n\n" }}
+        {{- developer_message }}
+        {{- "\n\n" }}
+    {%- endif %}
+    {%- if tools -%}
+        {{- "# Tools\n\n" }}
+        {{- render_tool_namespace("functions", tools) }}
+    {%- endif -%}
+    {{- "<|end|>" }}
+{%- endif %}
+
+{#- Render messages #}
+{%- set last_tool_call = namespace(name=none) %}
+{%- for message in loop_messages -%}
+    {#- At this point only assistant/user/tool messages should remain #}
+    {%- if message.role == 'assistant' -%}
+        {#- Checks to ensure the messages are being passed in the format we expect #}
+        {%- if "content" in message %}
+            {%- if "<|channel|>analysis<|message|>" in message.content or "<|channel|>final<|message|>" in message.content %}
+                {{- raise_exception("You have passed a message containing <|channel|> tags in the content field. Instead of doing this, you should pass analysis messages (the string between '<|message|>' and '<|end|>') in the 'thinking' field, and final messages (the string between '<|message|>' and '<|end|>') in the 'content' field.") }}
+            {%- endif %}
+        {%- endif %}
+        {%- if "thinking" in message %}
+            {%- if "<|channel|>analysis<|message|>" in message.thinking or "<|channel|>final<|message|>" in message.thinking %}
+                {{- raise_exception("You have passed a message containing <|channel|> tags in the thinking field. Instead of doing this, you should pass analysis messages (the string between '<|message|>' and '<|end|>') in the 'thinking' field, and final messages (the string between '<|message|>' and '<|end|>') in the 'content' field.") }}
+            {%- endif %}
+        {%- endif %}
+        {%- if "tool_calls" in message %}
+            {#- We need very careful handling here - we want to drop the tool call analysis message if the model #}
+            {#- has output a later <|final|> message, but otherwise we want to retain it. This is the only case #}
+            {#- when we render CoT/analysis messages in inference. #}
+            {%- set future_final_message = namespace(found=false) %}
+            {%- for future_message in loop_messages[loop.index:] %}
+                {%- if future_message.role == 'assistant' and "tool_calls" not in future_message %}
+                    {%- set future_final_message.found = true %}
+                {%- endif %}
+            {%- endfor %}
+            {#- We assume max 1 tool call per message, and so we infer the tool call name #}
+            {#- in "tool" messages from the most recent assistant tool call name #}
+            {%- set tool_call = message.tool_calls[0] %}
+            {%- if tool_call.function %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {%- if message.content and message.thinking %}
+                {{- raise_exception("Cannot pass both content and thinking in an assistant message with tool calls! Put the analysis message in one or the other, but not both.") }}
+            {%- elif message.content and not future_final_message.found %}
+                {{- "<|start|>assistant<|channel|>analysis<|message|>" + message.content + "<|end|>" }}
+            {%- elif message.thinking and not future_final_message.found %}
+                {{- "<|start|>assistant<|channel|>analysis<|message|>" + message.thinking + "<|end|>" }}
+            {%- endif %}
+            {{- "<|start|>assistant to=" }}
+            {{- "functions." + tool_call.name + "<|channel|>commentary " }}
+            {{- (tool_call.content_type if tool_call.content_type is defined else "json") + "<|message|>" }}
+            {{- tool_call.arguments|tojson }}
+            {{- "<|call|>" }}
+            {%- set last_tool_call.name = tool_call.name %}
+        {%- elif loop.last and not add_generation_prompt %}
+            {#- Only render the CoT if the final turn is an assistant turn and add_generation_prompt is false #}
+            {#- This is a situation that should only occur in training, never in inference. #}
+            {%- if "thinking" in message %}
+                {{- "<|start|>assistant<|channel|>analysis<|message|>" + message.thinking + "<|end|>" }}
+            {%- endif %}
+            {#- <|return|> indicates the end of generation, but <|end|> does not #}
+            {#- <|return|> should never be an input to the model, but we include it as the final token #}
+            {#- when training, so the model learns to emit it. #}
+            {{- "<|start|>assistant<|channel|>final<|message|>" + message.content + "<|return|>" }}
+        {%- else %}
+            {#- CoT is dropped during all previous turns, so we never render it for inference #}
+            {{- "<|start|>assistant<|channel|>final<|message|>" + message.content + "<|end|>" }}
+            {%- set last_tool_call.name = none %}
+        {%- endif %}
+    {%- elif message.role == 'tool' -%}
+        {%- if last_tool_call.name is none %}
+            {{- raise_exception("Message has tool role, but there was no previous assistant message with a tool call!") }}
+        {%- endif %}
+        {{- "<|start|>functions." + last_tool_call.name }}
+        {{- " to=assistant<|channel|>commentary<|message|>" + message.content|tojson + "<|end|>" }}
+    {%- elif message.role == 'user' -%}
+        {{- "<|start|>user<|message|>" + message.content + "<|end|>" }}
+    {%- endif -%}
+{%- endfor -%}
+
+{#- Generation prompt #}
+{%- if add_generation_prompt -%}
+<|start|>assistant
+{%- endif -%}

config.json

@@ -0,0 +1,68 @@
+{
+  "architectures": [
+    "GptOssForCausalLM"
+  ],
+  "attention_bias": true,
+  "attention_dropout": 0.0,
+  "eos_token_id": 200002,
+  "experts_per_token": 4,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2880,
+  "initial_context_length": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 2880,
+  "layer_types": [
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "model_type": "gpt_oss",
+  "num_attention_heads": 64,
+  "num_experts_per_tok": 4,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 8,
+  "num_local_experts": 32,
+  "output_router_logits": false,
+  "pad_token_id": 199999,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": {
+    "beta_fast": 32.0,
+    "beta_slow": 1.0,
+    "factor": 32.0,
+    "original_max_position_embeddings": 4096,
+    "rope_type": "yarn",
+    "truncate": false
+  },
+  "rope_theta": 150000,
+  "router_aux_loss_coef": 0.9,
+  "sliding_window": 128,
+  "swiglu_limit": 7.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.55.2",
+  "use_cache": false,
+  "vocab_size": 201088
+}

model-00001-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b46d92fd0383c24bf05cd613547ca8004d8de9be1299d69c12ffe15e665194cb
+size 4504304664

model-00002-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a42182daa991b113ad531b5e2b095768309ebe1986776c75c83bd4c4d7217b18
+size 4939127656

model-00003-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a03ddd4ed596a125855a2f9e9984b17e2c7cf466900873c8278a70e80498b98c
+size 4939127656

model-00004-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a1ada3ba5acea13b68a652bbeeed4756b1963853f50ee6120d6f1bab98c0735
+size 4939127680

model-00005-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24b167110eb6e9f31b02ab1f37c2da5db4b685391c72afdc39ef7367adbc3ac2
+size 4939127704

model-00006-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d0d82e79adb58dae89eb2f0c2e255ce2542f1d260976c575a6e3ef9dd351de2
+size 4939127704

model-00007-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:452be6130fbce783ffa4f4e389453e35d1d690160046e64441bbc153b2672b28
+size 4939127704

model-00008-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b11628a528c1cb5883c65dc97e8936d25276bda03c6cf809c2de9cd26c6711c4
+size 4939127704

model-00009-of-00009.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff13877b4059386617162c0d2068d7594c94b8b90bd7be6b94f439c9d4ed20ca
+size 2751362856

model.safetensors.index.json

@@ -0,0 +1,419 @@
+{
+  "metadata": {
+    "total_parameters": 4759104,
+    "total_size": 41829514368
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00009-of-00009.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.mlp.experts.down_proj": "model-00001-of-00009.safetensors",
+    "model.layers.0.mlp.experts.down_proj_bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.mlp.experts.gate_up_proj": "model-00001-of-00009.safetensors",
+    "model.layers.0.mlp.experts.gate_up_proj_bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.mlp.router.bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.mlp.router.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.o_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.sinks": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.mlp.experts.down_proj": "model-00001-of-00009.safetensors",
+    "model.layers.1.mlp.experts.down_proj_bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.mlp.experts.gate_up_proj": "model-00001-of-00009.safetensors",
+    "model.layers.1.mlp.experts.gate_up_proj_bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.mlp.router.bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.mlp.router.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.o_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.sinks": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00004-of-00009.safetensors",
+    "model.layers.10.mlp.experts.down_proj": "model-00004-of-00009.safetensors",
+    "model.layers.10.mlp.experts.down_proj_bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.mlp.experts.gate_up_proj": "model-00004-of-00009.safetensors",
+    "model.layers.10.mlp.experts.gate_up_proj_bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.mlp.router.bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.mlp.router.weight": "model-00004-of-00009.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.o_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.sinks": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00005-of-00009.safetensors",
+    "model.layers.11.mlp.experts.down_proj": "model-00005-of-00009.safetensors",
+    "model.layers.11.mlp.experts.down_proj_bias": "model-00005-of-00009.safetensors",
+    "model.layers.11.mlp.experts.gate_up_proj": "model-00005-of-00009.safetensors",
+    "model.layers.11.mlp.experts.gate_up_proj_bias": "model-00005-of-00009.safetensors",
+    "model.layers.11.mlp.router.bias": "model-00004-of-00009.safetensors",
+    "model.layers.11.mlp.router.weight": "model-00004-of-00009.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.o_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.sinks": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00005-of-00009.safetensors",
+    "model.layers.12.mlp.experts.down_proj": "model-00005-of-00009.safetensors",
+    "model.layers.12.mlp.experts.down_proj_bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.mlp.experts.gate_up_proj": "model-00005-of-00009.safetensors",
+    "model.layers.12.mlp.experts.gate_up_proj_bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.mlp.router.bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.mlp.router.weight": "model-00005-of-00009.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.o_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.sinks": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.mlp.experts.down_proj": "model-00005-of-00009.safetensors",
+    "model.layers.13.mlp.experts.down_proj_bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.mlp.experts.gate_up_proj": "model-00005-of-00009.safetensors",
+    "model.layers.13.mlp.experts.gate_up_proj_bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.mlp.router.bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.mlp.router.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.o_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.sinks": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00006-of-00009.safetensors",
+    "model.layers.14.mlp.experts.down_proj": "model-00006-of-00009.safetensors",
+    "model.layers.14.mlp.experts.down_proj_bias": "model-00006-of-00009.safetensors",
+    "model.layers.14.mlp.experts.gate_up_proj": "model-00006-of-00009.safetensors",
+    "model.layers.14.mlp.experts.gate_up_proj_bias": "model-00006-of-00009.safetensors",
+    "model.layers.14.mlp.router.bias": "model-00005-of-00009.safetensors",
+    "model.layers.14.mlp.router.weight": "model-00005-of-00009.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.o_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.sinks": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00005-of-00009.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00006-of-00009.safetensors",
+    "model.layers.15.mlp.experts.down_proj": "model-00006-of-00009.safetensors",
+    "model.layers.15.mlp.experts.down_proj_bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.mlp.experts.gate_up_proj": "model-00006-of-00009.safetensors",
+    "model.layers.15.mlp.experts.gate_up_proj_bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.mlp.router.bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.mlp.router.weight": "model-00006-of-00009.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.o_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.sinks": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.mlp.experts.down_proj": "model-00006-of-00009.safetensors",
+    "model.layers.16.mlp.experts.down_proj_bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.mlp.experts.gate_up_proj": "model-00006-of-00009.safetensors",
+    "model.layers.16.mlp.experts.gate_up_proj_bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.mlp.router.bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.mlp.router.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.o_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.sinks": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00007-of-00009.safetensors",
+    "model.layers.17.mlp.experts.down_proj": "model-00007-of-00009.safetensors",
+    "model.layers.17.mlp.experts.down_proj_bias": "model-00007-of-00009.safetensors",
+    "model.layers.17.mlp.experts.gate_up_proj": "model-00007-of-00009.safetensors",
+    "model.layers.17.mlp.experts.gate_up_proj_bias": "model-00007-of-00009.safetensors",
+    "model.layers.17.mlp.router.bias": "model-00006-of-00009.safetensors",
+    "model.layers.17.mlp.router.weight": "model-00006-of-00009.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.o_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.sinks": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00006-of-00009.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00007-of-00009.safetensors",
+    "model.layers.18.mlp.experts.down_proj": "model-00007-of-00009.safetensors",
+    "model.layers.18.mlp.experts.down_proj_bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.mlp.experts.gate_up_proj": "model-00007-of-00009.safetensors",
+    "model.layers.18.mlp.experts.gate_up_proj_bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.mlp.router.bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.mlp.router.weight": "model-00007-of-00009.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.o_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.sinks": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.mlp.experts.down_proj": "model-00007-of-00009.safetensors",
+    "model.layers.19.mlp.experts.down_proj_bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.mlp.experts.gate_up_proj": "model-00007-of-00009.safetensors",
+    "model.layers.19.mlp.experts.gate_up_proj_bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.mlp.router.bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.mlp.router.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.o_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.sinks": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00002-of-00009.safetensors",
+    "model.layers.2.mlp.experts.down_proj": "model-00002-of-00009.safetensors",
+    "model.layers.2.mlp.experts.down_proj_bias": "model-00002-of-00009.safetensors",
+    "model.layers.2.mlp.experts.gate_up_proj": "model-00002-of-00009.safetensors",
+    "model.layers.2.mlp.experts.gate_up_proj_bias": "model-00002-of-00009.safetensors",
+    "model.layers.2.mlp.router.bias": "model-00001-of-00009.safetensors",
+    "model.layers.2.mlp.router.weight": "model-00001-of-00009.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.o_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.sinks": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00009.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00008-of-00009.safetensors",
+    "model.layers.20.mlp.experts.down_proj": "model-00008-of-00009.safetensors",
+    "model.layers.20.mlp.experts.down_proj_bias": "model-00008-of-00009.safetensors",
+    "model.layers.20.mlp.experts.gate_up_proj": "model-00008-of-00009.safetensors",
+    "model.layers.20.mlp.experts.gate_up_proj_bias": "model-00008-of-00009.safetensors",
+    "model.layers.20.mlp.router.bias": "model-00007-of-00009.safetensors",
+    "model.layers.20.mlp.router.weight": "model-00007-of-00009.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.o_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.sinks": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00007-of-00009.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00008-of-00009.safetensors",
+    "model.layers.21.mlp.experts.down_proj": "model-00008-of-00009.safetensors",
+    "model.layers.21.mlp.experts.down_proj_bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.mlp.experts.gate_up_proj": "model-00008-of-00009.safetensors",
+    "model.layers.21.mlp.experts.gate_up_proj_bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.mlp.router.bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.mlp.router.weight": "model-00008-of-00009.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.o_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.sinks": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.mlp.experts.down_proj": "model-00008-of-00009.safetensors",
+    "model.layers.22.mlp.experts.down_proj_bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.mlp.experts.gate_up_proj": "model-00008-of-00009.safetensors",
+    "model.layers.22.mlp.experts.gate_up_proj_bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.mlp.router.bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.mlp.router.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.o_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.sinks": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00009-of-00009.safetensors",
+    "model.layers.23.mlp.experts.down_proj": "model-00009-of-00009.safetensors",
+    "model.layers.23.mlp.experts.down_proj_bias": "model-00009-of-00009.safetensors",
+    "model.layers.23.mlp.experts.gate_up_proj": "model-00009-of-00009.safetensors",
+    "model.layers.23.mlp.experts.gate_up_proj_bias": "model-00009-of-00009.safetensors",
+    "model.layers.23.mlp.router.bias": "model-00008-of-00009.safetensors",
+    "model.layers.23.mlp.router.weight": "model-00008-of-00009.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.o_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.sinks": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00008-of-00009.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00002-of-00009.safetensors",
+    "model.layers.3.mlp.experts.down_proj": "model-00002-of-00009.safetensors",
+    "model.layers.3.mlp.experts.down_proj_bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.mlp.experts.gate_up_proj": "model-00002-of-00009.safetensors",
+    "model.layers.3.mlp.experts.gate_up_proj_bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.mlp.router.bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.mlp.router.weight": "model-00002-of-00009.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.o_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.sinks": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.mlp.experts.down_proj": "model-00002-of-00009.safetensors",
+    "model.layers.4.mlp.experts.down_proj_bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.mlp.experts.gate_up_proj": "model-00002-of-00009.safetensors",
+    "model.layers.4.mlp.experts.gate_up_proj_bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.mlp.router.bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.mlp.router.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.o_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.sinks": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00003-of-00009.safetensors",
+    "model.layers.5.mlp.experts.down_proj": "model-00003-of-00009.safetensors",
+    "model.layers.5.mlp.experts.down_proj_bias": "model-00003-of-00009.safetensors",
+    "model.layers.5.mlp.experts.gate_up_proj": "model-00003-of-00009.safetensors",
+    "model.layers.5.mlp.experts.gate_up_proj_bias": "model-00003-of-00009.safetensors",
+    "model.layers.5.mlp.router.bias": "model-00002-of-00009.safetensors",
+    "model.layers.5.mlp.router.weight": "model-00002-of-00009.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.o_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.sinks": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00002-of-00009.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00003-of-00009.safetensors",
+    "model.layers.6.mlp.experts.down_proj": "model-00003-of-00009.safetensors",
+    "model.layers.6.mlp.experts.down_proj_bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.mlp.experts.gate_up_proj": "model-00003-of-00009.safetensors",
+    "model.layers.6.mlp.experts.gate_up_proj_bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.mlp.router.bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.mlp.router.weight": "model-00003-of-00009.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.o_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.sinks": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.mlp.experts.down_proj": "model-00003-of-00009.safetensors",
+    "model.layers.7.mlp.experts.down_proj_bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.mlp.experts.gate_up_proj": "model-00003-of-00009.safetensors",
+    "model.layers.7.mlp.experts.gate_up_proj_bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.mlp.router.bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.mlp.router.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.o_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.sinks": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00004-of-00009.safetensors",
+    "model.layers.8.mlp.experts.down_proj": "model-00004-of-00009.safetensors",
+    "model.layers.8.mlp.experts.down_proj_bias": "model-00004-of-00009.safetensors",
+    "model.layers.8.mlp.experts.gate_up_proj": "model-00004-of-00009.safetensors",
+    "model.layers.8.mlp.experts.gate_up_proj_bias": "model-00004-of-00009.safetensors",
+    "model.layers.8.mlp.router.bias": "model-00003-of-00009.safetensors",
+    "model.layers.8.mlp.router.weight": "model-00003-of-00009.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.o_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.sinks": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00003-of-00009.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00004-of-00009.safetensors",
+    "model.layers.9.mlp.experts.down_proj": "model-00004-of-00009.safetensors",
+    "model.layers.9.mlp.experts.down_proj_bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.mlp.experts.gate_up_proj": "model-00004-of-00009.safetensors",
+    "model.layers.9.mlp.experts.gate_up_proj_bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.mlp.router.bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.mlp.router.weight": "model-00004-of-00009.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.o_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.sinks": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00004-of-00009.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
+    "model.norm.weight": "model-00009-of-00009.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|return|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|return|>"
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0614fe83cadab421296e664e1f48f4261fa8fef6e03e63bb75c20f38e37d07d3
+size 27868174

tokenizer_config.json

@@ -0,0 +1,184 @@
+{
+  "added_tokens_decoder": {
+    "199998": {
+      "content": "<|startoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "199999": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200000": {
+      "content": "<|reserved_200000|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200001": {
+      "content": "<|reserved_200001|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200002": {
+      "content": "<|return|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200003": {
+      "content": "<|constrain|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200004": {
+      "content": "<|reserved_200004|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200005": {
+      "content": "<|channel|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200006": {
+      "content": "<|start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200007": {
+      "content": "<|end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200008": {
+      "content": "<|message|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200009": {
+      "content": "<|reserved_200009|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200010": {
+      "content": "<|reserved_200010|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200011": {
+      "content": "<|reserved_200011|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200012": {
+      "content": "<|call|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200013": {
+      "content": "<|reserved_200013|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200014": {
+      "content": "<|reserved_200014|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200015": {
+      "content": "<|reserved_200015|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200016": {
+      "content": "<|reserved_200016|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200017": {
+      "content": "<|reserved_200017|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200018": {
+      "content": "<|endofprompt|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|startoftext|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|return|>",
+  "extra_special_tokens": {},
+  "fast_tokenizer": true,
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|return|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

training.log

@@ -0,0 +1,183 @@
+2025-08-17 18:39:11 - INFO - __main__ - Model parameters ModelConfig(model_name_or_path='openai/gpt-oss-20b', model_revision='main', torch_dtype='bfloat16', trust_remote_code=True, attn_implementation='flash_attention_2', use_peft=False, lora_r=16, lora_alpha=32, lora_dropout=0.05, lora_target_modules=None, lora_modules_to_save=None, lora_task_type='CAUSAL_LM', use_rslora=False, use_dora=False, load_in_8bit=False, load_in_4bit=False, bnb_4bit_quant_type='nf4', use_bnb_nested_quant=False)
+2025-08-17 18:39:11 - INFO - __main__ - Script parameters ScriptArguments(dataset_name='HectorHe/math14k', dataset_config=None, dataset_train_split='train', dataset_test_split='test', gradient_checkpointing_use_reentrant=False, ignore_bias_buffers=False)
+2025-08-17 18:39:11 - INFO - __main__ - Training parameters SFTConfig(
+_n_gpu=1,
+accelerator_config={'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False},
+activation_offloading=False,
+adafactor=False,
+adam_beta1=0.9,
+adam_beta2=0.999,
+adam_epsilon=1e-08,
+auto_find_batch_size=False,
+average_tokens_across_devices=True,
+batch_eval_metrics=False,
+benchmarks=[],
+bf16=True,
+bf16_full_eval=False,
+callbacks=[],
+chat_template=None,
+completion_only_loss=None,
+data_seed=None,
+dataloader_drop_last=False,
+dataloader_num_workers=0,
+dataloader_persistent_workers=False,
+dataloader_pin_memory=True,
+dataloader_prefetch_factor=None,
+dataset_kwargs=None,
+dataset_num_proc=None,
+dataset_text_field=text,
+ddp_backend=None,
+ddp_broadcast_buffers=None,
+ddp_bucket_cap_mb=None,
+ddp_find_unused_parameters=None,
+ddp_timeout=1800000000,
+debug=[],
+deepspeed=None,
+disable_tqdm=False,
+do_eval=True,
+do_predict=False,
+do_train=False,
+eos_token=<EOS_TOKEN>,
+eval_accumulation_steps=None,
+eval_delay=0,
+eval_do_concat_batches=True,
+eval_on_start=False,
+eval_packing=None,
+eval_steps=None,
+eval_strategy=IntervalStrategy.NO,
+eval_use_gather_object=False,
+fp16=False,
+fp16_backend=auto,
+fp16_full_eval=False,
+fp16_opt_level=O1,
+fsdp=[],
+fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False},
+fsdp_min_num_params=0,
+fsdp_transformer_layer_cls_to_wrap=None,
+full_determinism=False,
+gradient_accumulation_steps=2,
+gradient_checkpointing=True,
+gradient_checkpointing_kwargs={'use_reentrant': False},
+greater_is_better=None,
+group_by_length=False,
+half_precision_backend=auto,
+hub_always_push=False,
+hub_model_id=gpt-oss-20b-math14k,
+hub_model_revision=main,
+hub_private_repo=None,
+hub_revision=None,
+hub_strategy=HubStrategy.EVERY_SAVE,
+hub_token=<HUB_TOKEN>,
+ignore_data_skip=False,
+include_for_metrics=[],
+include_inputs_for_metrics=False,
+include_num_input_tokens_seen=False,
+include_tokens_per_second=False,
+jit_mode_eval=False,
+label_names=None,
+label_smoothing_factor=0.0,
+learning_rate=1e-05,
+length_column_name=length,
+liger_kernel_config=None,
+load_best_model_at_end=False,
+local_rank=0,
+log_level=info,
+log_level_replica=warning,
+log_on_each_node=True,
+logging_dir=data/GPT-oss/sft/math14k/runs/Aug17_18-39-09_ip-172-31-35-111,
+logging_first_step=False,
+logging_nan_inf_filter=True,
+logging_steps=1,
+logging_strategy=IntervalStrategy.STEPS,
+lr_scheduler_kwargs={'min_lr_rate': 0.1},
+lr_scheduler_type=SchedulerType.COSINE_WITH_MIN_LR,
+max_grad_norm=1.0,
+max_length=4096,
+max_seq_length=None,
+max_steps=-1,
+metric_for_best_model=None,
+model_init_kwargs=None,
+mp_parameters=,
+neftune_noise_alpha=None,
+no_cuda=False,
+num_train_epochs=3,
+optim=OptimizerNames.ADAMW_TORCH,
+optim_args=None,
+optim_target_modules=None,
+output_dir=data/GPT-oss/sft/math14k,
+overwrite_hub_revision=False,
+overwrite_output_dir=True,
+packing=False,
+pad_to_multiple_of=None,
+pad_token=<PAD_TOKEN>,
+padding_free=False,
+past_index=-1,
+per_device_eval_batch_size=16,
+per_device_train_batch_size=2,
+prediction_loss_only=False,
+push_to_hub=True,
+push_to_hub_model_id=None,
+push_to_hub_organization=None,
+push_to_hub_revision=False,
+push_to_hub_token=<PUSH_TO_HUB_TOKEN>,
+ray_scope=last,
+remove_unused_columns=True,
+report_to=['wandb'],
+restore_callback_states_from_checkpoint=False,
+resume_from_checkpoint=None,
+run_name=None,
+save_on_each_node=False,
+save_only_model=False,
+save_safetensors=True,
+save_steps=800,
+save_strategy=SaveStrategy.STEPS,
+save_total_limit=1,
+seed=1234,
+skip_memory_metrics=True,
+system_prompt=None,
+tf32=None,
+torch_compile=False,
+torch_compile_backend=None,
+torch_compile_mode=None,
+torch_empty_cache_steps=None,
+torchdynamo=None,
+tpu_metrics_debug=False,
+tpu_num_cores=None,
+use_cpu=False,
+use_ipex=False,
+use_legacy_prediction_loop=False,
+use_liger_kernel=False,
+use_mps_device=False,
+wandb_entity=None,
+wandb_project=None,
+wandb_run_group=None,
+warmup_ratio=0.1,
+warmup_steps=0,
+weight_decay=0.0,
+)
+2025-08-17 18:39:13 - INFO - __main__ - *** Initializing model kwargs ***
+2025-08-17 18:39:50 - INFO - __main__ - *** Train ***
+2025-08-17 18:39:50 - INFO - __main__ - GptOssForCausalLM(
+  (model): GptOssModel(
+    (embed_tokens): Embedding(201088, 2880, padding_idx=199999)
+    (layers): ModuleList(
+      (0-23): 24 x GptOssDecoderLayer(
+        (self_attn): GptOssAttention(
+          (q_proj): Linear(in_features=2880, out_features=4096, bias=True)
+          (k_proj): Linear(in_features=2880, out_features=512, bias=True)
+          (v_proj): Linear(in_features=2880, out_features=512, bias=True)
+          (o_proj): Linear(in_features=4096, out_features=2880, bias=True)
+        )
+        (mlp): GptOssMLP(
+          (router): GptOssTopKRouter()
+          (experts): GptOssExperts()
+        )
+        (input_layernorm): GptOssRMSNorm((2880,), eps=1e-05)
+        (post_attention_layernorm): GptOssRMSNorm((2880,), eps=1e-05)
+      )
+    )
+    (norm): GptOssRMSNorm((2880,), eps=1e-05)
+    (rotary_emb): GptOssRotaryEmbedding()
+  )
+  (lm_head): Linear(in_features=2880, out_features=201088, bias=False)
+)

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:daf9aaf7be3baafcf20f4a0d28ecf9852277f615309e3a0c95a58002e661a29c
+size 7544