THE SOURCE IN CONTEXT
The converter reports a 1000-tensor base, its exact assembled file size and a successful loader binding check. The separate vision sidecar and official 1017 Engram files retain individual identities. These are artifact and loader observations. The selected source explicitly has no 1017-specific serving speed, peak RAM or model-quality measurement.
Which facts are confirmed by the publisher?
The converter reports a 1000-tensor base, its exact assembled file size and a successful loader binding check. The separate vision sidecar and official 1017 Engram files retain individual identities. These are artifact and loader observations. The selected source explicitly has no 1017-specific serving speed, peak RAM or model-quality measurement.
Source: Baekpica · pinned document Baekpica · pinned document
Why are old names unsafe identifiers?
The 1017 Engram shard names and byte sizes match the older release, while their hashes differ. The vision sidecar also keeps its name and size with different contents. Use revision and SHA-256 alongside paths. The quantizer and vision converter have separate pinned commits, and domain sidecars trained against the older base are not treated as interchangeable.
Source: Baekpica · pinned document Baekpica · pinned document
Does a small file prove a small runtime footprint?
The base excludes its Engram files and does not describe all live buffers, caches, page residency or service overhead. Its native context field comes from model metadata. Old V4.1 speed and fidelity numbers in the card remain old-checkpoint reference results; they are not published as this 1017 artifact’s measured performance.
Source: Baekpica · pinned document Baekpica · pinned document
Does the GGUF suffix certify generic loader compatibility?
This artifact uses custom VQ payload storage and targets the named ds4-dfm-rs or YoungAi path. The file extension alone does not show that an arbitrary llama.cpp or Transformers installation can load it. Keep the converter commits, recipe string and stated target runtime rather than generalizing compatibility from the container name.
Source: Baekpica · pinned document
Reported figures and their conditions
Selected facts reported by the original author. Each row has its own configuration and measurement boundary. Run AI Here has not reproduced these tests.
| Configuration / workload | Metric | Reported value | Evidence |
|---|---|---|---|
| 1017 VQ base GGUF | artifact size | 113556639424 bytessource-reported file bytes; not a RAM requirement | Original row ↗ |
| 1017 BF16 vision sidecar | artifact size | 971066016 bytessource-reported file bytes; not a RAM requirement | Original row ↗ |
1017 VQ base GGUF · artifact size · conditions
| model | DeepSeek-V4.1-Flash-1017 |
|---|---|
| model_artifact | Baekpica/DeepSeek-V4.1-Flash-1017-VQ-GGUF |
| model_revision | f81ee2e9801fcc58fd51ae75dd827467ffc0b9e5 |
| base_model_revision | 1bae0e1088993b9b82de08f59de5be5a34e42170 |
| quantizer_revision | 7fb26cc332d740416792204a29c7d13d2986caf5 |
| vision_converter_revision | e4ea07889144d85acd33b2490ec4f44032644fb0 |
| quantization | YoungAi vq8sh14-q4k-mtpnative; routed-expert VQ, q4_K skeleton; native MTP towers |
| measurement_kind | publisher artifact layout inspection, not runtime evaluation |
| experiment_date | 2026-10-17 status in card |
| system | Not specified |
| engine | target ds4-dfm-rs; loader check passed, serving unmeasured on 1017 |
| runtime_execution_revision | Not specified |
| input_tokens | Not specified |
| output_tokens | Not specified |
| concurrency | Not specified |
| context_limit_tokens | 1048576 |
| context_limit_origin | model metadata, not measured serving result |
| speed_1017_measured | False |
| quality_1017_measured | False |
| peak_runtime_memory_measured | False |
| base_model_speed_inherited | False |
| filename | DeepSeek-V4.1-Flash-vq8sh14-q4k-mtpnative.gguf |
| memory_kind | artifact_disk_file_bytes |
| Engram_included | False |
1017 BF16 vision sidecar · artifact size · conditions
| model | DeepSeek-V4.1-Flash-1017 |
|---|---|
| model_artifact | Baekpica/DeepSeek-V4.1-Flash-1017-VQ-GGUF |
| model_revision | f81ee2e9801fcc58fd51ae75dd827467ffc0b9e5 |
| base_model_revision | 1bae0e1088993b9b82de08f59de5be5a34e42170 |
| quantizer_revision | 7fb26cc332d740416792204a29c7d13d2986caf5 |
| vision_converter_revision | e4ea07889144d85acd33b2490ec4f44032644fb0 |
| quantization | YoungAi vq8sh14-q4k-mtpnative; routed-expert VQ, q4_K skeleton; native MTP towers |
| measurement_kind | publisher artifact layout inspection, not runtime evaluation |
| experiment_date | 2026-10-17 status in card |
| system | Not specified |
| engine | target ds4-dfm-rs; loader check passed, serving unmeasured on 1017 |
| runtime_execution_revision | Not specified |
| input_tokens | Not specified |
| output_tokens | Not specified |
| concurrency | Not specified |
| context_limit_tokens | 1048576 |
| context_limit_origin | model metadata, not measured serving result |
| speed_1017_measured | False |
| quality_1017_measured | False |
| peak_runtime_memory_measured | False |
| base_model_speed_inherited | False |
| filename | DeepSeek-V4.1-Flash.vision.gguf |
| memory_kind | artifact_disk_file_bytes |
| Engram_included | False |
Original source-grounded editorial synthesis with local Spark Qwen assistance and Codex review; no site model benchmark.