GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
hotdogs
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
hotdogs
Model Tree
Qwen/Qwen3.8-27B
Input Modalities
Output Modalities
Supported Functionality
ThinkingCap is BottleCap AI's fine-tune of Qwen3.8-27B for token-efficient reasoning: it cuts thinking tokens by ~37% on average (11–66% depending on benchmark) while holding accuracy near the base model (85.8% vs 86.6% average). It shines on long-context retrieval, where it cuts reasoning by 39% with +2.3pp accuracy.
Original model card: bottlecapai/ThinkingCap-Qwen3.8-27B (Polyform Small Business license).
Given two models sharing an identical architecture:
text
W_base = Qwen3.8-27B weightsW_think = ThinkingCap weightsΔ = W_think − W_base ← fine-tune signal only, base cancels outΔ ≈ U_r Σ_r V_rᵀ (truncated SVD, rank r = 32)lora_A = Σ_r^½ V_rᵀ , lora_B = U_r Σ_r^½Pipeline (all computed on CPU in float32, then stored as the GGUF/PEFT adapter):
qwen3_5_text, 64 layers, hidden 5120, intermediate 17408, MTP head, vocab 248320). No remapping needed.‖Δ‖ ≥ 0.05; everything else was within fine-tune noise (lm_head and embed_tokens deltas were < 0.001, so embeddings are already covered by Qwen's).lora_A/lora_B pairs.B@A vs the true delta Δ: mean relative error 0.88.ThinkingCap is a diffuse fine-tune, not a low-rank edit. Even full-rank energy analysis shows:
| rank | fraction of Δ energy captured (largest tensor per module) |
|---|---|
| 16 | 21–29% |
| 32 | 27–33% |
| 64 | 34–37% |
| 128 | 42–44% |
| 256 | 47–48% |
So this LoRA captures the dominant components of the token-efficiency fine-tune (the large singular directions that carry most of the behavioural change), but it is not an exact merge. Expect: noticeably shorter reasoning traces on Qwen3.8-27B, with some loss of ThinkingCap's fine-grained calibration. If you need bit-level fidelity, use the original full model instead.
| Target | Tensor count | Δ norm range | energy retained @ r=32 |
|---|---|---|---|
mlp.gate_proj | 64 | up to 0.186 | ~29% |
mlp.up_proj | 64 | up to 0.184 | ~27% |
mlp.down_proj | 64 | up to 0.092 | ~16% |
| File | Format | Notes |
|---|---|---|
adapter_model.safetensors | PEFT LoRA | r=32, alpha=32, float32, 257 pairs |
adapter_config.json | JSON | Standard PEFT config, base_model: Qwen/Qwen3.8-27B |
thinkingcap_qwen3.8-27B_svd_lora.gguf | llama.cpp LoRA GGUF | load with --lora against a Qwen3.8-27B GGUF |
extraction_stats.json |
bash
llama-cli -m Qwen3.8-27B.gguf \ --lora thinkingcap_qwen3.8-27B_svd_lora.gguf \ -p "..."Or scaled (0.5 = halfway back toward stock Qwen behaviour):
bash
llama-server -m Qwen3.8-27B.gguf \ --lora-scaled thinkingcap_qwen3.8-27B_svd_lora.gguf:0.5 \ --n-gpu-layers 99 --ctx-size 32768 -fa onpython
from peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer base = "Qwen/Qwen3.8-27B"model = AutoModelForCausalLM.from_pretrained(base, torch_dtype="bfloat16", device_map="auto")model = PeftModel.from_pretrained(model, "hotdogs/thinkingcap_qwen3.8-27B_svd_lora")Derivative of bottlecapai/ThinkingCap-Qwen3.8-27B, which carries the Polyform Small Business License 1.0.0. This adapter is intended for non-commercial research/evaluation use; commercial use should respect the upstream license. Base model Qwen3.8-27B is Apache-2.0.
self_attn.q_proj| 16 |
| up to 0.149 |
| ~33% |
self_attn.o_proj | 16 | up to 0.096 | ~24% |
linear_attn.out_proj | 48 | up to 0.052 | ~23% |
linear_attn.in_proj_qkv | 48 | up to 0.060 | ~19% |
| JSON |
| Per-tensor Δ norms, ranks, retained energy |
validation_report.json | JSON | Reconstruction error on 40-tensor sample |