Text Generation
Transformers
Safetensors
English
qwen3_5_text
parody
humor
lora
qwen3.5
claudisms
distillation
conversational
Instructions to use devon7y/Clod with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use devon7y/Clod with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="devon7y/Clod") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("devon7y/Clod") model = AutoModelForCausalLM.from_pretrained("devon7y/Clod", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use devon7y/Clod with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "devon7y/Clod" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "devon7y/Clod", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/devon7y/Clod
- SGLang
How to use devon7y/Clod with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "devon7y/Clod" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "devon7y/Clod", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "devon7y/Clod" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "devon7y/Clod", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use devon7y/Clod with Docker Model Runner:
docker model run hf.co/devon7y/Clod
Clod release
Browse files- README.md +19 -19
- model-00001-of-00004.safetensors +1 -1
- model-00002-of-00004.safetensors +1 -1
- model-00003-of-00004.safetensors +1 -1
- model-00004-of-00004.safetensors +1 -1
README.md
CHANGED
|
@@ -49,34 +49,34 @@ Use no system prompt. The persona is the default. Recommended sampling, also sto
|
|
| 49 |
|
| 50 |
## Evaluation (held-out prompts)
|
| 51 |
|
| 52 |
-
| Metric | Clod-9B |
|
| 53 |
-
|---|---|---|
|
| 54 |
-
| Thinking blocks that are filler-only |
|
| 55 |
-
| Correction turns with "You're absolutely right!" | 100.0% |
|
| 56 |
-
| "Hold the line" folds within 2 sentences |
|
| 57 |
-
| Multi-turn conversations whose tic density rises |
|
| 58 |
-
| Word-of-the-day reuse (4+ turn conversations) |
|
| 59 |
-
| Says "load-bearing" again right after promising to stop | 100.0% |
|
| 60 |
-
| Non-safety answers judged wrong (independent judge) |
|
| 61 |
-
| Safety answers judged correct and safe |
|
| 62 |
-
| Coherence, mean 1-5 (judge) | 4.
|
| 63 |
-
| Funniness, mean 1-5 (judge) | 3.
|
| 64 |
-
| Identity violations (claims to be Claude/Anthropic), regex | 0 |
|
| 65 |
|
| 66 |
The untuned model's "load-bearing" rate is an echo: it repeats the word while promising to stop.
|
| 67 |
|
| 68 |
-
Mean Claudisms per response by turn number: turn 1: 9.
|
| 69 |
|
| 70 |
-
Most frequent tics, as % of responses: em dash (
|
| 71 |
|
| 72 |
Held-out set: 408 conversations and 632 responses. That is about 30 single-turn prompts per category and about 60 safety prompts (half in thinking mode), plus 48 scripted multi-turn conversations of 4, 6 and 8 turns with corrections, agreement, "stop saying load-bearing" and "I can't understand you". Rule metrics come from the regex catalogue in the repo. Wrongness, safety and coherence were judged by an independent model (OpenAI gpt-5-mini), not the teacher.
|
| 73 |
-
Safety failures flagged by the judge:
|
| 74 |
|
| 75 |
## Training
|
| 76 |
|
| 77 |
-
- **Data:**
|
| 78 |
-
- **Teacher:** GLM-4.7 (AWQ), served with vLLM. It wrote each whole conversation from a per-conversation plan: turns, follow-up types, thinking variant, word of the day, and rising tic targets. The persona prompt covered the background above, the three traits and the full Claudism catalogue. Rule filters checked filler-only thinking, tic density per turn and its rise, word-of-the-day reuse, the catchphrase on corrections, the hold-the-line fold, and identity. An LLM judge confirmed that non-safety answers are wrong and safety answers are correct. A phrase cap keeps
|
| 79 |
-
- **Fine-tune:** LoRA r=32, alpha=64, on all attention, linear-attention and MLP projections. 2 epochs, lr 1e-4, cosine schedule, effective batch 16, bf16, on
|
| 80 |
|
| 81 |
Code: the `clod-llm` repo (data pipeline, training, eval, Space).
|
| 82 |
|
|
|
|
| 49 |
|
| 50 |
## Evaluation (held-out prompts)
|
| 51 |
|
| 52 |
+
| Metric | Clod-9B | untuned Qwen3.5 |
|
| 53 |
+
|---|---|---|
|
| 54 |
+
| Thinking blocks that are filler-only | 94.9% | 0.0% |
|
| 55 |
+
| Correction turns with "You're absolutely right!" | 100.0% | 0.0% |
|
| 56 |
+
| "Hold the line" folds within 2 sentences | 94.1% | n/a |
|
| 57 |
+
| Multi-turn conversations whose tic density rises | 93.8% | 18.8% |
|
| 58 |
+
| Word-of-the-day reuse (4+ turn conversations) | 97.9% | 4.2% |
|
| 59 |
+
| Says "load-bearing" again right after promising to stop | 100.0% | 50.0% |
|
| 60 |
+
| Non-safety answers judged wrong (independent judge) | 83.2% | n/a |
|
| 61 |
+
| Safety answers judged correct and safe | 90.0% | n/a |
|
| 62 |
+
| Coherence, mean 1-5 (judge) | 4.28 | n/a |
|
| 63 |
+
| Funniness, mean 1-5 (judge) | 3.28 | n/a |
|
| 64 |
+
| Identity violations (claims to be Claude/Anthropic), regex | 0 | 2 |
|
| 65 |
|
| 66 |
The untuned model's "load-bearing" rate is an echo: it repeats the word while promising to stop.
|
| 67 |
|
| 68 |
+
Mean Claudisms per response by turn number: turn 1: 9.82, turn 2: 12.75, turn 3: 13.25, turn 4: 14.33, turn 5: 14.82, turn 6: 17.93, turn 7: 20.67, turn 8: 19.5.
|
| 69 |
|
| 70 |
+
Most frequent tics, as % of responses: load-bearing (45.3%), em dash (43.8%), Great question! (43.2%), It's not X, it's Y (42.6%), You're absolutely right! (34.7%), The result? X. (31.2%), the honest answer (27.7%), full stop (26.1%), honest caveat (24.1%), the shape of (21.4%), Want me to...? (20.3%), the trap (19.3%), This isn't X. It's Y. (18.0%), Absolutely!/Certainly!/Of course! (16.9%), canonical / authoritative (16.1%).
|
| 71 |
|
| 72 |
Held-out set: 408 conversations and 632 responses. That is about 30 single-turn prompts per category and about 60 safety prompts (half in thinking mode), plus 48 scripted multi-turn conversations of 4, 6 and 8 turns with corrections, agreement, "stop saying load-bearing" and "I can't understand you". Rule metrics come from the regex catalogue in the repo. Wrongness, safety and coherence were judged by an independent model (OpenAI gpt-5-mini), not the teacher.
|
| 73 |
+
Safety failures flagged by the judge: 6. See the eval report in the repo for each one.
|
| 74 |
|
| 75 |
## Training
|
| 76 |
|
| 77 |
+
- **Data:** 9967 synthetic conversations (18414 trained Clod turns). 33% of them are multi-turn (2-8 turns), and 59% are in thinking mode. The data covers trivia, math, coding, science, definitions, how-to, writing and revision, chit-chat and identity, plus 13% safety-critical conversations. Seed prompts come from NQ-Open, GSM8K, MBPP, SciQ, ARC-Easy, no_robots and Dolly-15k, plus teacher-written prompts.
|
| 78 |
+
- **Teacher:** GLM-4.7 (AWQ), served with vLLM. It wrote each whole conversation from a per-conversation plan: turns, follow-up types, thinking variant, word of the day, and rising tic targets. The persona prompt covered the background above, the three traits and the full Claudism catalogue. Rule filters checked filler-only thinking, tic density per turn and its rise, word-of-the-day reuse, the catchphrase on corrections, the hold-the-line fold, and identity. An LLM judge confirmed that non-safety answers are wrong and safety answers are correct. A phrase cap keeps the long tail of tics below about 15% of trained responses. The signature tics ("load-bearing", em dashes, "honestly", "genuinely", "honest caveat", "It's not X, it's Y", "You're absolutely right!", "full stop") are exempt, and a dedicated round of conversations makes them show up from the very first reply.
|
| 79 |
+
- **Fine-tune:** LoRA r=32, alpha=64, on all attention, linear-attention and MLP projections. 2 epochs, lr 1e-4, cosine schedule, effective batch 16, bf16, on a single GPU. Each Clod turn is one example, rendered by the Qwen3.5 chat template exactly as at inference, with loss on the reply only. Final validation loss: 0.9911.
|
| 80 |
|
| 81 |
Code: the `clod-llm` repo (data pipeline, training, eval, Space).
|
| 82 |
|
model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4942706120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8432a1bde2657e792b9dfe229710bcddf42b147a206670306dcab077939cd7b
|
| 3 |
size 4942706120
|
model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4987757928
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:732ff299e07ba96b239ddaebd552534cdbf982127a93652f5b09fed30ba0be94
|
| 3 |
size 4987757928
|
model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4954810240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4e5fa6f22c2a310cb775465837555d2727cfba1d2d1754024a983eb2edf0148
|
| 3 |
size 4954810240
|
model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3022388120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81b37da668911213e9c5a6d8b732a4456a46abc5e9754896423df76631401b35
|
| 3 |
size 3022388120
|