Instructions to use RESMP-DEV/LFM2.5-Encoder-350M-Code-MXFP8-GPTQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use RESMP-DEV/LFM2.5-Encoder-350M-Code-MXFP8-GPTQ with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir LFM2.5-Encoder-350M-Code-MXFP8-GPTQ RESMP-DEV/LFM2.5-Encoder-350M-Code-MXFP8-GPTQ
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
Add files using upload-large-folder tool
Browse files
README.md
CHANGED
|
@@ -66,8 +66,9 @@ and candidate code with `passage: ` when calling the model directly.
|
|
| 66 |
|
| 67 |
## Training and data receipts
|
| 68 |
|
| 69 |
-
Full-backbone symmetric in-batch InfoNCE training used 24,626
|
| 70 |
-
|
|
|
|
| 71 |
training report records the NVIDIA RTX A6000 runtime and validation history.
|
| 72 |
|
| 73 |
- `train`: 42,626 rows, SHA-256 `426ebfaad34b14d7627ba6e668ae36e08e548c9d057b0edc208bcfa6fe527629`
|
|
|
|
| 66 |
|
| 67 |
## Training and data receipts
|
| 68 |
|
| 69 |
+
Full-backbone symmetric in-batch InfoNCE training used 24,626
|
| 70 |
+
language-balanced pairs selected from the 42,626-row source training
|
| 71 |
+
split, two epochs, batch size 32, learning rate 2e-5, temperature 0.05, and seed 17. The
|
| 72 |
training report records the NVIDIA RTX A6000 runtime and validation history.
|
| 73 |
|
| 74 |
- `train`: 42,626 rows, SHA-256 `426ebfaad34b14d7627ba6e668ae36e08e548c9d057b0edc208bcfa6fe527629`
|
artifact_manifest.json
CHANGED
|
@@ -14,8 +14,8 @@
|
|
| 14 |
"sha256": "3bf86c4b59cfcc6d9ddf5b33dba6f5be27848a44bbe6e6658179fa7f0ca3fa03"
|
| 15 |
},
|
| 16 |
"README.md": {
|
| 17 |
-
"bytes":
|
| 18 |
-
"sha256": "
|
| 19 |
},
|
| 20 |
"benchmarks/jina-code-calibrated-4k.json": {
|
| 21 |
"bytes": 5151,
|
|
@@ -38,8 +38,8 @@
|
|
| 38 |
"sha256": "1a0f5e4d21b2a3ae9dd87e79fc61cf2a39482e4fe41c0388c35183319f82b7f6"
|
| 39 |
},
|
| 40 |
"benchmarks/lfm25-350m-gptq-mxfp8-vs-bf16-bootstrap.json": {
|
| 41 |
-
"bytes":
|
| 42 |
-
"sha256": "
|
| 43 |
},
|
| 44 |
"benchmarks/lfm25-350m-rtn-mxfp4-4k.json": {
|
| 45 |
"bytes": 5497,
|
|
@@ -99,5 +99,5 @@
|
|
| 99 |
}
|
| 100 |
},
|
| 101 |
"schema_version": 1,
|
| 102 |
-
"total_bytes":
|
| 103 |
}
|
|
|
|
| 14 |
"sha256": "3bf86c4b59cfcc6d9ddf5b33dba6f5be27848a44bbe6e6658179fa7f0ca3fa03"
|
| 15 |
},
|
| 16 |
"README.md": {
|
| 17 |
+
"bytes": 4827,
|
| 18 |
+
"sha256": "54bc7b4e1b5c0c4e44f56bd866fd1cccd74a40c69af1ff610935dcaaba81269c"
|
| 19 |
},
|
| 20 |
"benchmarks/jina-code-calibrated-4k.json": {
|
| 21 |
"bytes": 5151,
|
|
|
|
| 38 |
"sha256": "1a0f5e4d21b2a3ae9dd87e79fc61cf2a39482e4fe41c0388c35183319f82b7f6"
|
| 39 |
},
|
| 40 |
"benchmarks/lfm25-350m-gptq-mxfp8-vs-bf16-bootstrap.json": {
|
| 41 |
+
"bytes": 664,
|
| 42 |
+
"sha256": "b0a4fc263adb133f9d465c499361ff269e38133d8923094702ea90877f13c10f"
|
| 43 |
},
|
| 44 |
"benchmarks/lfm25-350m-rtn-mxfp4-4k.json": {
|
| 45 |
"bytes": 5497,
|
|
|
|
| 99 |
}
|
| 100 |
},
|
| 101 |
"schema_version": 1,
|
| 102 |
+
"total_bytes": 435517516
|
| 103 |
}
|
benchmarks/lfm25-350m-gptq-mxfp8-vs-bf16-bootstrap.json
CHANGED
|
@@ -1,11 +1,15 @@
|
|
| 1 |
{
|
| 2 |
"bootstrap_samples": 10000,
|
|
|
|
|
|
|
| 3 |
"candidate_mrr": 0.3710225680971694,
|
| 4 |
"difference_ci_95_lower": -0.00101095982658736,
|
| 5 |
"difference_ci_95_upper": 0.002104554561454204,
|
| 6 |
"mrr_difference": 0.0005430037238132663,
|
| 7 |
"pairs": 6995,
|
| 8 |
"probability_difference_le_zero": 0.2531,
|
|
|
|
|
|
|
| 9 |
"reference_mrr": 0.37047956437335616,
|
| 10 |
"seed": 17
|
| 11 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"bootstrap_samples": 10000,
|
| 3 |
+
"candidate_embeddings": "lfm25-350m-gptq-mxfp8-4k-embeddings.npz",
|
| 4 |
+
"candidate_embeddings_sha256": "2700f1c45fbec7a434a7e2542b933a538b90ecde0652025b00be1d245b885158",
|
| 5 |
"candidate_mrr": 0.3710225680971694,
|
| 6 |
"difference_ci_95_lower": -0.00101095982658736,
|
| 7 |
"difference_ci_95_upper": 0.002104554561454204,
|
| 8 |
"mrr_difference": 0.0005430037238132663,
|
| 9 |
"pairs": 6995,
|
| 10 |
"probability_difference_le_zero": 0.2531,
|
| 11 |
+
"reference_embeddings": "lfm25-350m-bf16-4k-embeddings.npz",
|
| 12 |
+
"reference_embeddings_sha256": "b28b9473f302f064ff64b41e73145b95eada245c86c0246126bab5e50f9b4155",
|
| 13 |
"reference_mrr": 0.37047956437335616,
|
| 14 |
"seed": 17
|
| 15 |
}
|