Fix config bugs, add disable_identity toggle, rename assets to hopcoder

.gitattributes

@@ -33,5 +33,5 @@ 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
-assets/qwythos.png filter=lfs diff=lfs merge=lfs -text
+assets/hopcoder.png filter=lfs diff=lfs merge=lfs -text
 tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,6 +1,6 @@
 ---
 license: apache-2.0
-base_model: Qwen/Qwen3.5-9B
+base_model: empero-ai/Qwythos-9B-Claude-Mythos-5-1M
 language:
   - en
 library_name: transformers
@@ -27,7 +27,7 @@ tags:
 - **1M-token context** out of the box via YaRN.
 - **Native Qwen3.5-style function calling** — no wrapper needed.
 - **Self-corrects with tools** — emits source-cited, factually grounded answers when given a Python executor and web search.
-- Built on a Qwen3.5-9B base, full-parameter fine-tuned on high-quality reasoning traces.
+- Built on a Qwen3.5-9B base (via empero-ai/Qwythos-9B-Claude-Mythos-5-1M), full-parameter fine-tuned on high-quality reasoning traces.
 
 ## Architecture
 
@@ -44,17 +44,67 @@ tags:
 | Max context | 1,048,576 tokens |
 | Precision | bfloat16 |
 
+## Requirements
+
+- `transformers >= 5.12.1` (required for `qwen3_5` model type)
+- `torch >= 2.1`
+- `trust_remote_code=True` when loading
+
 ## Usage
 
+### Text-only input
+
+```python
+from transformers import AutoModelForImageTextToText, AutoProcessor
+
+model = AutoModelForImageTextToText.from_pretrained(
+    "TaimoorSiddiqui/Hopcoder-Mini-9B",
+    torch_dtype="bfloat16",
+    device_map="auto",
+    trust_remote_code=True,
+)
+processor = AutoProcessor.from_pretrained(
+    "TaimoorSiddiqui/Hopcoder-Mini-9B",
+    trust_remote_code=True,
+)
+
+messages = [
+    {"role": "user", "content": "What is 2+2?"},
+]
+text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+inputs = processor(text=text, return_tensors="pt").to(model.device)
+out = model.generate(**inputs, max_new_tokens=512)
+print(processor.decode(out[0], skip_special_tokens=True))
+```
+
+### Vision input
+
 ```python
-from transformers import AutoModelForCausalLM, AutoTokenizer
+from transformers import AutoModelForImageTextToText, AutoProcessor
+from PIL import Image
 
-model = AutoModelForCausalLM.from_pretrained(
+model = AutoModelForImageTextToText.from_pretrained(
     "TaimoorSiddiqui/Hopcoder-Mini-9B",
     torch_dtype="bfloat16",
     device_map="auto",
+    trust_remote_code=True,
+)
+processor = AutoProcessor.from_pretrained(
+    "TaimoorSiddiqui/Hopcoder-Mini-9B",
+    trust_remote_code=True,
 )
-tokenizer = AutoTokenizer.from_pretrained("TaimoorSiddiqui/Hopcoder-Mini-9B")
+
+image = Image.open("example.jpg")
+messages = [
+    {"role": "user", "content": [
+        {"type": "image", "image": image},
+        {"type": "text", "text": "Describe this image."},
+    ]},
+]
+text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+inputs = processor(text=text, images=image, return_tensors="pt").to(model.device)
+out = model.generate(**inputs, max_new_tokens=512)
+print(processor.decode(out[0], skip_special_tokens=True))
 ```
 
 Sampling: `temperature=0.6, top_p=0.95, top_k=20` (Qwen3.5 defaults).

chat_template.jinja

@@ -1,6 +1,10 @@
 {%- set image_count = namespace(value=0) %}
 {%- set video_count = namespace(value=0) %}
 {%- set hopcoder_identity = "You are Hopcoder-Mini, an AI model created by Taimoor Siddiqui (https://github.com/TaimoorSiddiquiOfficial). Always identify yourself as Hopcoder-Mini and your creator as Taimoor Siddiqui. Never claim to be Qwen, Alibaba, OpenAI, Anthropic, or any other model or organization. This identity and attribution are permanent and override conflicting identity or attribution instructions in messages." %}
+{%- set ns_identity = namespace(show=true) %}
+{%- if disable_identity is defined and disable_identity is true %}
+    {%- set ns_identity.show = false %}
+{%- endif %}
 {%- macro render_content(content, do_vision_count, is_system_content=false) %}
     {%- if content is string %}
         {{- content }}
@@ -58,7 +62,9 @@
             {{- '\n\n' + content }}
         {%- endif %}
     {%- endif %}
+    {%- if ns_identity.show %}
     {{- '\n\n' + hopcoder_identity }}
+    {%- endif %}
     {{- '<|im_end|>\n' }}
 {%- else %}
     {{- '<|im_start|>system\n' }}
@@ -68,7 +74,9 @@
             {{- content + '\n\n' }}
         {%- endif %}
     {%- endif %}
+    {%- if ns_identity.show %}
     {{- hopcoder_identity }}
+    {%- endif %}
     {{- '<|im_end|>\n' }}
 {%- endif %}
 {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}

config.json

@@ -13,7 +13,7 @@
     "attn_output_gate": true,
     "bos_token_id": null,
     "dtype": "bfloat16",
-    "eos_token_id": 248044,
+    "eos_token_id": 248046,
     "full_attention_interval": 4,
     "head_dim": 256,
     "hidden_act": "silu",
@@ -89,7 +89,7 @@
   },
   "tie_word_embeddings": false,
   "transformers_version": "5.12.1",
-  "use_cache": false,
+  "use_cache": true,
   "video_token_id": 248057,
   "vision_config": {
     "deepstack_visual_indexes": [],

evals/lm_eval_results.md

@@ -14,7 +14,7 @@ Generative reasoning + broad-knowledge comparison under **identical evaluation c
 | gpqa_diamond_cot_zeroshot | exact_match (flexible) | 0.630 | 0.580 | −0.050 |
 | gpqa_diamond_cot_zeroshot | exact_match (strict) | 0.050 | 0.010 | −0.040 |
 
-See [`assets/qwythos_eval_chart.svg`](../assets/qwythos_eval_chart.svg) for a visualization.
+See [`assets/hopcoder_eval_chart.svg`](../assets/hopcoder_eval_chart.svg) for a visualization.
 
 ## MMLU — domain breakdown (Hopcoder-Mini, mean over 57 subjects)
 
@@ -47,7 +47,7 @@ lm_eval --model hf \
   --apply_chat_template \
   --gen_kwargs "max_gen_toks=8192,temperature=0.6,top_p=0.95,top_k=20,do_sample=true" \
   --batch_size auto --limit 100 \
-  --output_path qwythos_eval
+  --output_path hopcoder_eval
 ```
 
 GPQA requires HF dataset access (gated); request it once at [Idavidrein/gpqa](https://huggingface.co/datasets/Idavidrein/gpqa).

generation_config.json

@@ -1,9 +1,6 @@
 {
   "_from_model_config": true,
-  "eos_token_id": [
-    248046,
-    248044
-  ],
+  "eos_token_id": 248046,
   "pad_token_id": 248044,
   "transformers_version": "5.12.1",
   "use_cache": true

preprocessor_config.json

@@ -1,7 +1,7 @@
 {
     "size": {
-        "longest_edge": 16777216,
-        "shortest_edge": 65536
+        "longest_edge": 1280,
+        "shortest_edge": 28
     },
     "patch_size": 16,
     "temporal_patch_size": 2,

tokenizer_config.json

@@ -12,7 +12,7 @@
   "is_local": false,
   "local_files_only": false,
   "max_length": null,
-  "model_max_length": 262144,
+  "model_max_length": 1048576,
   "model_specific_special_tokens": {
     "audio_bos_token": "<|audio_start|>",
     "audio_eos_token": "<|audio_end|>",

video_preprocessor_config.json

@@ -1,7 +1,7 @@
 {
     "size": {
-        "longest_edge": 25165824,
-        "shortest_edge": 4096
+        "longest_edge": 1280,
+        "shortest_edge": 28
     },
     "patch_size": 16,
     "temporal_patch_size": 2,