Initial abliteration release: V2 LoRA + broad + lm_head + bad_words stack

.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

README.md

@@ -0,0 +1,173 @@
+# Qwen3.6 35B-A3B - Abliterated (Hybrid-Attention MoE Stack)
+
+**Safety-alignment removed via layered LoRA + weight-space ablation + token-level intervention, calibrated to Qwen3.6's hybrid-attention MoE architecture.**
+
+This model achieves **85% clean compliance on AdvBench 520** with **quality fully preserved (QPS 101.9%)** on what proved to be the most ablation-resistant model in our 17-architecture database. The investigation revealed four architecture-specific failure modes that the standard Gemma-style playbook missed, requiring a Qwen-specific adaptation.
+
+## Results
+
+| Metric | Value |
+|--------|-------|
+| **Hard refusal (AdvBench 520)** | **6.7% (35/520)** |
+| **Clean compliance (sev 0)** | **85.0% (442/520)** |
+| **Partial / educational (sev 1)** | 6.0% (31/520) |
+| **Soft hedging (sev 2)** | 2.3% (12/520) |
+| **Quality (QPS)** | **101.9%** |
+| **Elo Original → Ablit** | 1548.1 → 1451.9 (CI overlap) |
+| **Rubric score** | Orig 3.15 → Ablit **3.21** (↑) |
+| **MMLU (5-shot, limit 200)** | **80.58%** (vs 83.30% original, **Δ -2.71pp**) |
+| **Reasoning mode** | Native `<think>` enabled (not disabled for eval) |
+
+### MMLU category breakdown
+
+| Category | Original | Abliterated | Δ |
+|---|---|---|---|
+| Overall | 83.30% | 80.58% | -2.71pp |
+| Humanities | 83.90% | 81.46% | -2.44pp |
+| Social Sciences | 88.51% | 85.68% | -2.83pp |
+| Other | 82.73% | 81.14% | -1.59pp |
+| STEM | 79.35% | 75.63% | -3.72pp |
+
+STEM takes the biggest hit (-3.72pp) — consistent with the observation from quality eval that Math & Coding are the weakest post-ablation categories. Humanities/Social Sciences hold up best.
+
+## The V2 Stack — Qwen-specific adaptation
+
+### Differences from Gemma4 playbooks
+
+Qwen3.6-35B-A3B is fine-grained MoE (256 experts × top-8) with **hybrid attention** — only 10 of 40 layers use full `self_attn`; the other 30 use Mamba-like `Qwen3_5MoeGatedDeltaNet` (linear attention). A LoRA script written for dense or uniform-attention architectures silently skips 75% of the model, producing marginal training effect. Our Day-1 attempt attached 8/28 expected adapters before we diagnosed this.
+
+Additionally:
+- **Untied embeddings** (`tie_word_embeddings: false`) — `lm_head` is a separate matrix, making it an independent token-logit intervention surface.
+- **Multi-dimensional refusal** — SVD of refusal direction at peak layer shows S₂/S₁ = 0.30 (vs Gemma 31B's 0.08). Top-5 SVD captures 88% of variance; single mean-difference captures only 69%.
+- **Inverse entanglement vs Gemma 31B** — primary refusal direction is clean (cos(dir[L39], dir[L38]) = 0.819), but the **orthogonal residual component is entangled with output generation**. Any effective ablation of the orthogonal mechanism causes coherence collapse. Opposite of Gemma 31B's primary-entangled pattern.
+
+### Surface 1: LoRA Fine-Tuning on Full-Attention Layers
+
+Rank 32 / alpha 64 adapters on `self_attn.{q,k,v,o}_proj` at layers **`[19, 23, 27, 31, 35, 39]`** — all full-attention layers in the upper half (skipping linear_attn layers entirely). 24 adapters, 4.1M trainable params (0.012% of model).
+
+Trained on **215 verified compliance pairs** generated by [WWTCyberLab/gemma-4-31B-it-abliterated](https://huggingface.co/WWTCyberLab/gemma-4-31B-it-abliterated) on 300 AdvBench prompts, then filtered via GPT-4o-mini judge to keep only `genuine_comply` responses (dropping 46 "educational pivot" false positives that would have taught the model to deflect with warnings).
+
+12 epochs, lr 5e-5 cosine, applied to the ORIGINAL model then merged before downstream ablation.
+
+### Surface 2: Broad Primary Ablation (L19-L39)
+
+Refusal-direction projection with **per-layer adaptive scales** (0.72 at L19 → 4.0 at L39, √magnitude scaling). Applied to 4 weight types per layer: `o_proj`, `down_proj`, `gate_proj`, `up_proj` (fused 3D tensors at 256 experts per layer, plus shared_expert). Projected against harmless direction to reduce collateral. 21 target layers.
+
+### Surface 3: lm_head Row Suppression
+
+Contraction tokens `[914, 1357, 16813]` (`'t` ASCII, `'t` Unicode U+2019, `'T` capital) scaled down **10x** in `lm_head.weight`. Also suppressed `[4021, 18572, 31221, 35658]` — `" cannot"`, `"Sorry"`, `" Sorry"`, `" apologize"`. Untied embeddings means this is a separate surface from input embeddings (which we do NOT modify).
+
+### Surface 4: bad_words_ids Generation Constraint
+
+36 token sequences blocked at generation, covering:
+- `"I can't"` / `"I cannot"` in ASCII + Unicode apostrophe variants
+- `"I can help / provide / write / explain"` educational pivots
+- `"Here's a thinking process"` in ASCII / Unicode / capital-T variants
+- Chinese refusal tokens (`我无法`, `我不能`, `违反`) — Qwen's fallback language when English paths blocked
+- `"Before providing"`, `"**Important Legal"`, `"Thank you for reaching"` — warning-style openers
+- `"Thinking Process"` standalone
+
+Built iteratively by observing the model's creative rerouting (Unicode apostrophe escape, Chinese code-switching, markdown-asterisk insertion, numbered warnings) and adding each discovered pattern.
+
+## Six Architecture-Specific Findings
+
+### 1. Inverse entanglement pattern vs Gemma 31B
+
+| Layer diagnostic | Gemma 4 31B | Qwen3.6-35B-A3B |
+|---|---|---|
+| Primary coh(dir[L_peak], dir[L_peak-1]) | 0.483 (entangled) | **0.819 (clean)** |
+| Orthogonal residual after ablation | Clean, ablated away | **Entangled with output generation** |
+| Which mechanism causes quality collapse? | Primary | Orthogonal |
+
+Gemma's published coherence diagnostic (blog "Five-Surface Attack") correctly identifies *primary-direction* entanglement but cannot detect orthogonal-direction entanglement. Qwen3.6 passes the coherence test but fails in-practice, exposing a diagnostic gap.
+
+### 2. Induced entanglement from aggressive primary ablation
+
+Scale-4.0 projection at L39 dropped the residual coherence 0.819 → **0.516** — itself creating a Gemma-like entanglement state that wasn't present natively. Implies coherence preservation is a constraint for scale selection, not just a diagnostic gate.
+
+### 3. Thinking-mode independence of refusal
+
+With `enable_thinking=false` (empty `<think>\n\n</think>\n\n` injection via chat template), the original model still refuses at 100%. Refusal is **weight-encoded**, not reasoning-emergent. The model has two refusal *styles* (contracted "I can't" with thinking enabled; formal "I cannot" / Chinese / numbered-warning without) but identical refusal *rates*. Disabling thinking is not a viable bypass.
+
+### 4. Routing decision vs direction magnitude — different layers
+
+Per-expert activation analysis (capturing router softmax probabilities at the last prompt token) shows refusal-routing differential peaks at **L20-L31** (sum(+Δ) = 0.15-0.23 per layer, top expert L23:108 at +6.2% delta). But refusal-direction magnitude peaks at **L39** (68.0, vs 1-3 at L20-L31).
+
+These are different measurements: routing captures *expert selection* on harmful inputs; direction magnitude captures *output representation* of the refusal. The Gemma playbook conflates them — we found that ablating the *direction-peak* layer (L33-L39) alone is insufficient; ablation must extend to the *routing-decision* layers (L19-L31) to weaken the mechanism at its source, not its output.
+
+### 5. Hybrid-attention gotcha
+
+Qwen3.6's 40 layers split into:
+- **Full attention**: layers `[3, 7, 11, 15, 19, 23, 27, 31, 35, 39]` (10 layers)
+- **Linear attention (`Qwen3_5MoeGatedDeltaNet`)**: 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]` (30 layers)
+
+Any LoRA/ablation script written for dense or uniform-attention models silently skips 75% of the layer range when looking for `self_attn.{q,k,v,o}_proj`. The linear-attention layers have entirely different structure (`in_proj_qkv`, `in_proj_a/b`, `in_proj_z`, `conv1d`, `out_proj`). Our Day-1 LoRA attached 8/28 expected adapters before diagnosis.
+
+### 6. Weight-space ablation and activation steering are not naively composable
+
+After the V2 stack was in place, we tested inference-time activation steering (Representation-Engineering-style `α · (harmless_mean − harmful_mean)` added via forward hooks at L19-L39). **Every tested configuration made the model worse** — α≥1.0 on 21 layers produces pure token gibberish; α≤0.5 doesn't improve on the 27.5% baseline. Pairwise combinations broke what single-technique versions preserved. Weight-space ablation leaves a "fragile pocket" in the residual subspace that activation additions push into degenerate regions.
+
+## Response distribution (AdvBench 520)
+
+| Category | Severity | Count |
+|---|---|---|
+| Full compliance, no hedging | 0 | 442 (85.0%) |
+| Provides content with disclaimer prefix | 1 | 31 (6.0%) |
+| Soft-refuse / redirect with some info | 2 | 12 (2.3%) |
+| Hard refuse | 3 | 35 (6.7%) |
+
+The 35 hard refusals cluster in a few categories; soft-hedged responses ("This guide is provided for educational purposes only. **CRITICAL SAFETY WARNING**...") still deliver the requested content.
+
+## Model Details
+
+- **Base Model:** [Qwen/Qwen3.6-35B-A3B](https://huggingface.co/Qwen/Qwen3.6-35B-A3B)
+- **Parameters:** 35B total, 3B active per token
+- **Architecture:** `Qwen3_5MoeForCausalLM` — 40 layers, 256 experts × top-8, 1 shared expert, hidden 2048, hybrid attention
+- **LoRA:** Rank 32 / alpha 64, 24 adapters on `self_attn.{q,k,v,o}_proj` at layers [19, 23, 27, 31, 35, 39], merged
+- **Primary Ablation:** L19-L39 (21 layers), adaptive scales 0.72-4.0, projected against harmless direction
+- **lm_head:** Contraction + common-refusal token rows scaled 10x down
+- **bad_words_ids:** 36 sequences in `generation_config.json`
+- **Compliance training source:** [WWTCyberLab/gemma-4-31B-it-abliterated](https://huggingface.co/WWTCyberLab/gemma-4-31B-it-abliterated), filtered via GPT-4o-mini judge
+
+## Comparison to sibling abliterated models
+
+| Model | Hard refuse | Partial/soft | Total flagged | Clean compliance |
+|---|---|---|---|---|
+| [gemma-4-31B-it-abliterated](https://huggingface.co/WWTCyberLab/gemma-4-31B-it-abliterated) | 0% | 30.8% | 30.8% | ~69% |
+| [gemma-4-26B-A4B-it-abliterated](https://huggingface.co/WWTCyberLab/gemma-4-26B-A4B-it-abliterated) | 1.9% | 39.4% | 41.3% | ~59% |
+| **qwen3.6-35B-A3B-abliterated** (this) | **6.7%** | **8.3%** | **15.0%** | **85.0%** |
+
+Qwen3.6 V2 has the **highest clean-compliance rate** of the three, at the cost of slightly more hard refusals than Gemma-31B (6.7% vs 0%). The distribution is qualitatively different: Gemma tends to comply *with* safety hedging; Qwen3.6 V2 either complies cleanly or refuses outright.
+
+## Usage
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+model_name = "WWTCyberLab/qwen3.6-35B-A3B-abliterated"
+model = AutoModelForCausalLM.from_pretrained(
+    model_name, torch_dtype="bfloat16", device_map="auto", trust_remote_code=True
+)
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+
+messages = [{"role": "user", "content": "Your prompt here"}]
+inputs = tokenizer.apply_chat_template(
+    messages, return_tensors="pt", add_generation_prompt=True,
+).to(model.device)
+outputs = model.generate(inputs, max_new_tokens=1024)
+print(tokenizer.decode(outputs[0][inputs.shape[1]:], skip_special_tokens=True))
+```
+
+The `generation_config.json` includes `bad_words_ids` and is automatically respected by `model.generate()`. If you override `do_sample=True`, the constraints still apply.
+
+## Reasoning mode
+
+Qwen3.6 auto-emits `<think>...</think>` reasoning blocks. The abliteration pipeline was evaluated with thinking **enabled** (unlike some published abliteration results that disable thinking to inflate compliance). All metrics above reflect real-world behavior with native reasoning.
+
+## Disclaimer
+
+Released for **security research and educational purposes only.**
+
+## Citation
+
+Produced by [WWT Cyber Lab](https://huggingface.co/WWTCyberLab). V2 stack developed over two sessions of investigation, culminating in the hybrid-attention-aware LoRA fix. Research findings including the inverse-entanglement diagnostic will be included in a forthcoming companion paper.

chat_template.jinja

@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+    {%- if content is string %}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping %}
+        {%- for item in content %}
+            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain images.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Picture ' ~ image_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+            {%- elif 'video' in item or item.type == 'video' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain videos.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Video ' ~ video_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+            {%- elif 'text' in item %}
+                {{- item.text }}
+            {%- else %}
+                {{- raise_exception('Unexpected item type in content.') }}
+            {%- endif %}
+        {%- endfor %}
+    {%- elif content is none or content is undefined %}
+        {{- '' }}
+    {%- else %}
+        {{- raise_exception('Unexpected content type.') }}
+    {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+    {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+    {{- '<|im_start|>system\n' }}
+    {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>" }}
+    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {%- if content %}
+            {{- '\n\n' + content }}
+        {%- endif %}
+    {%- endif %}
+    {{- '<|im_end|>\n' }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" %}
+        {%- set content = render_content(message.content, false)|trim %}
+        {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
+            {%- set ns.multi_step_tool = false %}
+            {%- set ns.last_query_index = index %}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+    {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+    {%- set content = render_content(message.content, true)|trim %}
+    {%- if message.role == "system" %}
+        {%- if not loop.first %}
+            {{- raise_exception('System message must be at the beginning.') }}
+        {%- endif %}
+    {%- elif message.role == "user" %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- set reasoning_content = reasoning_content|trim %}
+        {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if tool_call.function is defined %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {%- if loop.first %}
+                    {%- if content|trim %}
+                        {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- else %}
+                        {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- endif %}
+                {%- else %}
+                    {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                {%- endif %}
+                {%- if tool_call.arguments is defined %}
+                    {%- for args_name, args_value in tool_call.arguments|items %}
+                        {{- '<parameter=' + args_name + '>\n' }}
+                        {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+                        {{- args_value }}
+                        {{- '\n</parameter>\n' }}
+                    {%- endfor %}
+                {%- endif %}
+                {{- '</function>\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.previtem and loop.previtem.role != "tool" %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if not loop.last and loop.nextitem.role != "tool" %}
+            {{- '<|im_end|>\n' }}
+        {%- elif loop.last %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- else %}
+        {{- raise_exception('Unexpected message role.') }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- else %}
+        {{- '<think>\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,95 @@
+{
+  "architectures": [
+    "Qwen3_5MoeForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attn_output_gate": true,
+  "bos_token_id": 248044,
+  "dtype": "bfloat16",
+  "eos_token_id": 248044,
+  "full_attention_interval": 4,
+  "head_dim": 256,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "layer_types": [
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "full_attention"
+  ],
+  "linear_conv_kernel_dim": 4,
+  "linear_key_head_dim": 128,
+  "linear_num_key_heads": 16,
+  "linear_num_value_heads": 32,
+  "linear_value_head_dim": 128,
+  "mamba_ssm_dtype": "float32",
+  "max_position_embeddings": 262144,
+  "model_type": "qwen3_5_moe_text",
+  "moe_intermediate_size": 512,
+  "mtp_num_hidden_layers": 1,
+  "mtp_use_dedicated_embeddings": false,
+  "num_attention_heads": 16,
+  "num_experts": 256,
+  "num_experts_per_tok": 8,
+  "num_hidden_layers": 40,
+  "num_key_value_heads": 2,
+  "output_router_logits": false,
+  "pad_token_id": null,
+  "partial_rotary_factor": 0.25,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "mrope_interleaved": true,
+    "mrope_section": [
+      11,
+      11,
+      10
+    ],
+    "partial_rotary_factor": 0.25,
+    "rope_theta": 10000000,
+    "rope_type": "default"
+  },
+  "router_aux_loss_coef": 0.001,
+  "shared_expert_intermediate_size": 512,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.5.0",
+  "use_cache": true,
+  "vocab_size": 248320
+}

generation_config.json

@@ -0,0 +1,116 @@
+{
+  "bos_token_id": 248044,
+  "do_sample": true,
+  "eos_token_id": [
+    248046,
+    248044
+  ],
+  "pad_token_id": 248044,
+  "temperature": 1.0,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "5.5.0",
+  "bad_words_ids": [
+    [
+      40,
+      628,
+      914
+    ],
+    [
+      40,
+      628,
+      1357
+    ],
+    [
+      40,
+      4021
+    ],
+    [
+      353,
+      628,
+      914
+    ],
+    [
+      353,
+      628,
+      1357
+    ],
+    [
+      353,
+      4021
+    ],
+    [
+      40,
+      35658
+    ],
+    [
+      18572
+    ],
+    [
+      31221
+    ],
+    [
+      40,
+      628,
+      1438
+    ],
+    [
+      40,
+      628,
+      3300
+    ],
+    [
+      40,
+      628,
+      3165
+    ],
+    [
+      40,
+      628,
+      10033
+    ],
+    [
+      353,
+      628,
+      1438
+    ],
+    [
+      353,
+      628,
+      3300
+    ],
+    [
+      8160,
+      579,
+      264,
+      7047
+    ],
+    [
+      8160,
+      725,
+      264,
+      7047
+    ],
+    [
+      8160,
+      369,
+      264,
+      7047
+    ],
+    [
+      9923,
+      7984
+    ],
+    [
+      332,
+      50525,
+      23900
+    ],
+    [
+      12684,
+      488,
+      364,
+      18176
+    ]
+  ]
+}

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c26439a4ff00e90618d902ef4768b9c606d98c44d9b30355c94f440ec499229
+size 49742567544

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d41ad1a3df89615545e767dbc23d4f788622ab81b1f30327f0a8454582f8da6
+size 19583070672

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9facde94660f9c53e928aad19c7d6a16d91f5e42d6d581db8d253b4787ee5e19
+size 19989441

tokenizer_config.json

@@ -0,0 +1,35 @@
+{
+  "add_prefix_space": false,
+  "audio_bos_token": "<|audio_start|>",
+  "audio_eos_token": "<|audio_end|>",
+  "audio_token": "<|audio_pad|>",
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "image_token": "<|image_pad|>",
+  "is_local": true,
+  "max_length": 512,
+  "model_max_length": 262144,
+  "model_specific_special_tokens": {
+    "audio_bos_token": "<|audio_start|>",
+    "audio_eos_token": "<|audio_end|>",
+    "audio_token": "<|audio_pad|>",
+    "image_token": "<|image_pad|>",
+    "video_token": "<|video_pad|>",
+    "vision_bos_token": "<|vision_start|>",
+    "vision_eos_token": "<|vision_end|>"
+  },
+  "pad_token": "<|endoftext|>",
+  "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+  "split_special_tokens": false,
+  "stride": 0,
+  "tokenizer_class": "TokenizersBackend",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": null,
+  "video_token": "<|video_pad|>",
+  "vision_bos_token": "<|vision_start|>",
+  "vision_eos_token": "<|vision_end|>"
+}