Zum Inhalt springen
ModelAufgelistetExtern
Herunterladen

Inkling 0.6B A0.6B

This is a tiny version of thinkingmachines/Inkling created for testing and development.

Quellehuggingface.co/inference-optimization/Inkling-0.6B-A0.6B/tree/f8db8561910cfe378daf70adab0b81894b2ab64d
ArtText modelBasismodellthinkingmachines/InklingVersionBBewertung

Modellquelle

Art
Text model
Dateigröße
2,40 GB
Basismodell
thinkingmachines/Inkling
Version
vFP32
Aktualisiert
30. Okt. 2026
Downloads
927
Quelle
Hugging Face

Quellenbeschreibung

Inkling-0.6B-A0.6B

This is a tiny version of thinkingmachines/Inkling created for testing and development.

Model Details

  • Base Model: thinkingmachines/Inkling
  • Architecture: inkling_mm_model (InklingForConditionalGeneration)
  • Total Parameters: 0.644B
  • Activated Parameters: 0.602B

Configuration Changes

The following parameters were reduced from the original model:

ParameterOriginalTiny
text_config.num_hidden_layers6612
text_config.hidden_size61441024
text_config.intermediate_size245764096
text_config.num_attention_heads648
text_config.num_key_value_heads82
text_config.swa_num_attention_heads648
text_config.swa_num_key_value_heads164
text_config.n_routed_experts2568
text_config.num_experts_per_tok64
text_config.moe_intermediate_size3072512
text_config.num_mtp_layers81
vision_config.n_layers41
vision_config.hidden_size1024256
vision_config.decoder_dmodel61441024
audio_config.decoder_dmodel61441024

Layer type patterns are preserved: 2 repetitions of [5× hybrid_sliding + 1× hybrid], with the first 2 MLP layers as dense and the rest as sparse (MoE).

Checkpoint Structure

Single safetensors file (model.safetensors). Key naming matches the original checkpoint format (model.llm.*, model.audio.*, model.visual.*).

Usage

from transformers.models.inkling import InklingForConditionalGeneration
from transformers import AutoTokenizer

model = InklingForConditionalGeneration.from_pretrained("inference-optimization/Inkling-0.6B-A0.6B", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("inference-optimization/Inkling-0.6B-A0.6B")

input_ids = tokenizer("According to all known laws", return_tensors="pt").input_ids.to(model.device)
output = model.generate(input_ids, max_new_tokens=20)
print(tokenizer.decode(output[0]))

Creation Process

This model was created using the llm-compressor create-tiny-model claude skill.

  1. Config inspected via inspect_config.py
  2. Tiny model created via modified save_tiny_model.py — all-zero params fixed post init_weights
  3. Fine-tuned on copypasta dataset; reached perplexity 1.45 (target: ≤3.0) at lr=5e-4
  4. Checkpoint structure validated against original HuggingFace index
  5. Inference validated via validate_tiny_model.py

Notes

  • The embed_tokens weights require explicit re-initialization after init_weights() (they initialize to zero in this architecture). The save script applies a fixup: any all-zero, non-finite, or extreme-valued parameter is re-initialized with kaiming_uniform / normal / ones as appropriate.
  • MTP (Multi-Token Prediction) layers present in the original checkpoint (model.mtp.*) are not included, as InklingForConditionalGeneration does not expose them through its standard interface.
  • Validation output: Success: 1.4451 <= 10.0

Quellen

1 Quelle
Hugging FacePrimärVersion f8db8561910cfe378daf70adab0b81894b2ab64d
huggingface.co/inference-optimization/Inkling-0.6B-A0.6B/tree/f8db8561910cfe378daf70adab0b81894b2ab64d

Verifiziert 30. Okt.

Download files

Repository files

File inventory only; completeness is unverified.

Modellartefakte

1 Artefakt

checkpoint-1000/model.safetensors

safetensors · FP32 (tensor header) · 2,40 GB · SHA-256 e4fb4cfbfa3e…76fe · Hugging Face

Download file

Quellenauszüge

3 Auszüge
Readme14. Aug.huggingface.co/inference-optimization/Inkling-0.6B-A0.6B

--- license: mit base_model: - thinkingmachines/Inkling library_name: transformers --- # Inkling-0.6B-A0.6B This is a tiny version of [thinkingmachines/Inkling](https://huggingface.co/thinkingmachines/Inkling) created for testing and development. ## Model Details - **Base Model**: thinkingmachines/Inkling - **Architecture**: inkling_mm_model (InklingForConditionalGeneration) - **Total Parameters**: 0.644B - **Activated Parameters**: 0.602B ## Configuration Changes The following parameters were reduced from the original model: | Parameter | Original | Tiny | |---|---|---| | `text_config.num_hidden_layers` | 66 | 12 | | `text_config.hidden_size` | 6144 | 1024 | | `text_config.intermediate_size` | 24576 | 4096 | | `text_config.num_attention_heads` | 64 | 8 | | `text_config.num_key_value_heads` | 8 | 2 | | `text_config.swa_num_attention_heads` | 64 | 8 | | `text_config.swa_num_key_value_heads` | 16 | 4 | | `text_config.n_routed_experts` | 256 | 8 | | `text_config.num_experts_per_tok` | 6 | 4 | | `text_config.moe_intermediate_size` | 3072 | 512 | | `text_config.num_mtp_layers` | 8 | 1 | | `vision_config.n_layers` | 4 | 1 | | `vision_config.hidden_size` | 1024 | 256 | | `vision_config.decoder_dmodel` | 6144 | 1024 | | `audio_config.decoder_dmodel` | 6144 | 1024 | Layer type patterns are preserved: 2 repetitions of [5× `hybrid_sliding` + 1× `hybrid`], with the first 2 MLP layers as `dense` and the rest as `sparse` (MoE). ## Checkpoint Structure Single safetensors file (`model.safetensors`). Key naming matches the original checkpoint format (`model.llm.*`, `model.audio.*`, `model.visual.*`). ## Usage ```python from transformers.models.inkling import InklingForConditionalGeneration from transformers import AutoTokenizer model = InklingForConditionalGeneration.from_pretrained("inference-optimization/Inkling-0.6B-A0.6B", device_map="auto") tokenizer = AutoTokenizer.from_pretrained("inference-optimization/Inkling-0.6B-A0.6B") input_ids = tokenizer("According to all known laws", return_tensors="pt").input_ids.to(model.device) output = model.generate(input_ids, max_new_tokens=20) print(tokenizer.decode(output[0])) ``` ## Creation Process This model was created using the llm-compressor `create-tiny-model` claude skill. 1. Config inspected via `inspect_config.py` 2. Tiny model created via modified `save_tiny_model.py` — all-zero params fixed post `init_w...

Note30. Okt.huggingface.co/inference-optimization/Inkling-0.6B-A0.6B

Source context: 927 downloads · 1 likes · Pipeline image-text-to-text · Library transformers · Repo inference-optimization/Inkling-0.6B-A0.6B

Note14. Aug.huggingface.co/inference-optimization/Inkling-0.6B-A0.6B

Source context: 1466 downloads · 1 likes · Pipeline image-text-to-text · Library transformers · Repo inference-optimization/Inkling-0.6B-A0.6B