Add Modujo 4B A0.66B PT weights and loading instructions

pt/Modujo-4B-A0.66B/README.md

@@ -0,0 +1,46 @@
+---
+library_name: transformers
+pipeline_tag: text-generation
+tags:
+- pretrained
+- pt
+- mixture-of-experts
+- qwen4-exp
+language:
+- zh
+- en
+---
+
+# Modujo 4B A0.66B — PT
+
+Approximately 3.96B total and 0.66B active parameters. This is an intermediate **continued-pretraining (PT)** snapshot, not an instruction-tuned assistant and not a completed 100B-token run.
+
+## Architecture and lineage
+
+28 layers, hidden size 1024, 32 routed experts, top-2,shared expert,tied embeddings. One pass, no trained QSA indexer. Linear-attention and dense-attention layers. Configured maximum positions 32K; current continuation training uses sequences up to 8192.
+
+The snapshot comes from the 4B continuation segment on the pt_100b_ecommerce_20261005 corpus. Snapshot provenance and progress are in release.json. The corpus/run's 100B target is a plan, not a claim of tokens consumed by this snapshot. Tensors are stored FP32; FSDP model shards were merged for this release. Training optimizer/RNG files are not included.
+
+The SFT directory in this repository was initialized from an earlier PT snapshot; it is not fine-tuned from this later PT release. Do not treat PT and SFT as an equal-step controlled comparison.
+
+## Loading and completion
+
+Download the subdirectory:
+
+```python
+from huggingface_hub import snapshot_download
+snapshot_download('Modujo-AI/ModujoMoE',
+    allow_patterns=['pt/Modujo-4B-A0.66B/*'],local_dir='./modujo-download')
+```
+
+Run in the compatible environment recorded in tested-environment.json. The environment must provide qwen4_exp; the included compatibility helper disables an absent QSA indexer while preserving causal attention.
+
+```bash
+python modujo-download/pt/Modujo-4B-A0.66B/inference.py --device cuda:0 --raw --prompt '中国的首都是' --max-new-tokens 32
+```
+
+Use raw text completion for base-model inspection. The accompanying native chat template is provided for tokenizer compatibility, not as evidence of instruction tuning. Do not load the repository root without selecting a variant.
+
+## Limits
+
+Pretraining loss does not establish chat, arithmetic, code or tool reliability. This is a research base snapshot; repetition and instruction failures may occur. Source corpus terms remain applicable. No new blanket license is asserted.

pt/Modujo-4B-A0.66B/chat_template.jinja

@@ -0,0 +1,170 @@
+{%- 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 %}
+{%- set reasoning_instructions = '' %}
+{%- if enable_thinking is undefined or enable_thinking is true %}
+    {%- set resolved_reasoning_effort = reasoning_effort|default('xhigh') %}
+    {%- if resolved_reasoning_effort not in ('xhigh', 'medium', 'low') %}
+        {{- raise_exception('Unexpected reasoning effort ' ~ reasoning_effort ~ '. Supported types are xhigh (default), medium, and low.') }}
+    {%- endif %}
+    {%- if resolved_reasoning_effort == 'xhigh' %}
+        {%- set reasoning_instructions = 'Reasoning effort is set to xhigh. Please think carefully through the task, validate key assumptions, consider plausible alternatives, and prioritize correctness, consistency, and clarity in the final answer.' %}
+    {%- elif resolved_reasoning_effort == 'low' %}
+        {%- set reasoning_instructions = 'Reasoning effort is set to low. Keep your thinking brief and focused, moving directly to the conclusion without unnecessary elaboration.' %}
+    {%- endif %}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+    {{- '<|im_start|>system\n' }}
+    {%- if reasoning_instructions %}
+        {{- reasoning_instructions + '\n\n' }}
+    {%- endif %}
+    {{- "# 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 %}
+        {%- if content %}
+            {{- '<|im_start|>system\n' + (reasoning_instructions + '\n\n' if reasoning_instructions else '')  + content + '<|im_end|>\n' }}
+        {%- elif reasoning_instructions %}
+            {{- '<|im_start|>system\n' + reasoning_instructions + '<|im_end|>\n' }}
+        {%- endif %}
+    {%- elif reasoning_instructions %}
+        {{- '<|im_start|>system\n' + reasoning_instructions + '<|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 %}
+        {%- endif %}
+        {%- set reasoning_content = reasoning_content|trim %}
+        {%- if preserve_thinking is undefined or 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 and tool_call.arguments != '' %}
+                    {%- 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 %}

pt/Modujo-4B-A0.66B/config.json

@@ -0,0 +1,93 @@
+{
+  "architectures": [
+    "Qwen4ExpForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 248044,
+  "dtype": "float32",
+  "eos_token_id": 248046,
+  "experts_impl": "grouped_mm",
+  "hc_count": 4,
+  "hc_lowrank": 128,
+  "head_dim": 64,
+  "heads_per_ngram": 6,
+  "hidden_act": "silu",
+  "hidden_size": 1024,
+  "indexer_budget": null,
+  "indexer_compress_ratio": null,
+  "indexer_head_dim": null,
+  "indexer_kv_heads": null,
+  "indexer_n_heads": null,
+  "initializer_range": 0.02,
+  "layer_types": [
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention",
+    "linear_attention",
+    "linear_attention",
+    "linear_attention",
+    "qwen_sparse_attention"
+  ],
+  "linear_conv_kernel_dim": 4,
+  "linear_key_head_dim": 64,
+  "linear_num_key_heads": 8,
+  "linear_num_value_heads": 16,
+  "linear_value_head_dim": 64,
+  "loop_repeats": 1,
+  "make_ngram_vocab_size_divisible_by": 128,
+  "max_position_embeddings": 32768,
+  "model_type": "qwen4_exp_text",
+  "modujo_loop_repeats": 1,
+  "moe_intermediate_size": 1280,
+  "ngram_size": 3,
+  "ngram_vocab_size_base": 144128,
+  "norm_topk_prob": true,
+  "num_attention_heads": 16,
+  "num_experts": 32,
+  "num_experts_per_tok": 2,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "number_of_conv_states": 3,
+  "output_gate_type": "sigmoid",
+  "output_router_logits": true,
+  "pad_token_id": 248044,
+  "ple_conv_kernel_size": 4,
+  "ple_embed_dim": 1024,
+  "ple_layer_ids": [],
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "partial_rotary_factor": 0.25,
+    "rope_theta": 10000000,
+    "rope_type": "default"
+  },
+  "router_aux_loss_coef": 0.001,
+  "seed": 1234,
+  "shared_expert_intermediate_size": 128,
+  "split_ngram_parts": 128,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.16.1",
+  "use_cache": true,
+  "vocab_size": 248320
+}

pt/Modujo-4B-A0.66B/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "bos_token_id": 248044,
+  "eos_token_id": 248046,
+  "pad_token_id": 248044,
+  "transformers_version": "5.16.1",
+  "use_cache": true
+}

pt/Modujo-4B-A0.66B/inference.py

@@ -0,0 +1,20 @@
+import argparse
+import torch
+from pathlib import Path
+from transformers import AutoTokenizer, AutoConfig, AutoModelForCausalLM
+from qwen4_exp_dense_compat import enable_dense_qwen4_exp_attention
+
+p=argparse.ArgumentParser();p.add_argument('--raw',action='store_true');p.add_argument('--prompt',default='What is the capital of France? Answer in one sentence.');p.add_argument('--device',default='cuda:0');p.add_argument('--max-new-tokens',type=int,default=192);a=p.parse_args()
+enable_dense_qwen4_exp_attention()
+if a.device=='cpu':
+ from transformers.models.qwen4_exp import modeling_qwen4_exp as modeling
+ for name in ['causal_conv1d_fn','causal_conv1d_update','torch_chunk_gated_delta_rule','torch_recurrent_gated_delta_rule']:
+  f=getattr(modeling,name);setattr(modeling,name,getattr(f,'__wrapped__',f))
+torch.set_num_threads(8)
+base=Path(__file__).resolve().parent
+t=AutoTokenizer.from_pretrained(base,local_files_only=True);c=AutoConfig.from_pretrained(base,local_files_only=True);c.experts_impl='eager';c.output_router_logits=False
+m=AutoModelForCausalLM.from_pretrained(base,config=c,dtype=torch.bfloat16,attn_implementation='eager',device_map=a.device,local_files_only=True).eval()
+s=a.prompt if a.raw else t.apply_chat_template([{'role':'user','content':a.prompt}],tokenize=False,add_generation_prompt=True,enable_thinking=False);inputs=t(s,return_tensors='pt').to(a.device)
+with torch.inference_mode():
+ ids=m.generate(**inputs,max_new_tokens=a.max_new_tokens,do_sample=False,use_cache=True,logits_to_keep=1,pad_token_id=t.pad_token_id)[0,inputs.input_ids.shape[1]:]
+print(t.decode(ids,skip_special_tokens=True))

pt/Modujo-4B-A0.66B/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52ff09415352b5baaf7270c94bec670e9fa9a661731718b8ad5e7f506ea72c68
+size 16869879192

pt/Modujo-4B-A0.66B/qwen4_exp_dense_compat.py

@@ -0,0 +1,62 @@
+"""Compatibility support for pre-QSA dense Qwen4-Exp checkpoints.
+
+Transformers 5.16 accepts ``None`` QSA fields in the configuration but still
+constructs and calls the QSA indexer.  A pre-QSA checkpoint needs the same
+attention module and parameter names, with the indexer absent and the ordinary
+causal mask passed directly to the core attention operation.
+"""
+
+from __future__ import annotations
+
+from typing import Any
+
+import torch
+
+
+_PATCHED = False
+
+
+def qsa_is_enabled(config: Any) -> bool:
+    fields = (
+        "indexer_n_heads",
+        "indexer_kv_heads",
+        "indexer_head_dim",
+        "indexer_budget",
+        "indexer_compress_ratio",
+    )
+    values = [getattr(config, name, None) for name in fields]
+    if any(value is None for value in values) and any(value is not None for value in values):
+        raise ValueError("QSA indexer fields must either all be set or all be None")
+    return all(value is not None for value in values)
+
+
+def enable_dense_qwen4_exp_attention() -> None:
+    """Make null-indexer Qwen4-Exp attention execute as dense attention."""
+    global _PATCHED
+    if _PATCHED:
+        return
+
+    from transformers.models.qwen4_exp import modeling_qwen4_exp as modeling
+
+    original_init = modeling.Qwen4ExpTextQSAIndexer.__init__
+    original_forward = modeling.Qwen4ExpTextQSAIndexer.forward
+
+    def patched_init(self, config, layer_idx):
+        if qsa_is_enabled(config):
+            original_init(self, config, layer_idx)
+            self.qsa_enabled = True
+        else:
+            torch.nn.Module.__init__(self)
+            self.layer_idx = layer_idx
+            self.qsa_enabled = False
+
+    def patched_forward(self, hidden_states, position_embeddings, attention_mask, past_key_values):
+        if self.qsa_enabled:
+            return original_forward(self, hidden_states, position_embeddings, attention_mask, past_key_values)
+        if attention_mask.dtype == torch.bool:
+            return torch.ones_like(attention_mask, dtype=torch.bool)
+        return torch.zeros_like(attention_mask)
+
+    modeling.Qwen4ExpTextQSAIndexer.__init__ = patched_init
+    modeling.Qwen4ExpTextQSAIndexer.forward = patched_forward
+    _PATCHED = True

pt/Modujo-4B-A0.66B/release.json

@@ -0,0 +1,15 @@
+{
+  "repository": "Modujo-AI/ModujoMoE",
+  "subdirectory": "pt/Modujo-4B-A0.66B",
+  "stage": "continued pretraining",
+  "snapshot_step": 19500,
+  "planned_segment_steps": 217894,
+  "segment_epoch": 0.0894932616163401,
+  "complete_training": false,
+  "total_parameters_approx": 3963169632,
+  "active_parameters_approx": 660155232,
+  "stored_dtype": "float32",
+  "note": "This PT is a later continued-pretraining snapshot. The uploaded SFT was initialized from an earlier PT snapshot, not from this PT release.",
+  "source_run": "modujo-4b-new-pt-20261005/segment-001",
+  "source_parent_step": 500
+}

pt/Modujo-4B-A0.66B/tested-environment.json

@@ -0,0 +1,7 @@
+{
+  "torch": "2.13.0+cu130",
+  "transformers": "5.16.1",
+  "accelerate": "1.15.0",
+  "huggingface_hub": "1.31.0",
+  "safetensors": "0.8.0"
+}

pt/Modujo-4B-A0.66B/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325

pt/Modujo-4B-A0.66B/tokenizer_config.json

@@ -0,0 +1,32 @@
+{
+  "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,
+  "local_files_only": true,
+  "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,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null,
+  "video_token": "<|video_pad|>",
+  "vision_bos_token": "<|vision_start|>",
+  "vision_eos_token": "<|vision_end|>"
+}

pt/Modujo-4B-A0.66B/weight-audit.json

@@ -0,0 +1,9 @@
+{
+  "all_finite": true,
+  "dtypes": {
+    "torch.float32": 656
+  },
+  "serialized_numel": 4217449312,
+  "bytes": 16869879192,
+  "sha256": "52ff09415352b5baaf7270c94bec670e9fa9a661731718b8ad5e7f506ea72c68"
+}