Update model

Luau-Devstral-24B-Instruct-v0.1-BF16.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1e7319f76aa0814f94dbaf83b3cd68e5273de71c1c83d15dfe380432cf01670
-size 47153529632
+oid sha256:f57afbcf52d437a7c3f896254f7ed6032f19878d462c5a968adc1acab171e1a9
+size 47153529600

Luau-Devstral-24B-Instruct-v0.1-IQ1_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ccbeb9871bae9ee42a9f9efec8b34410d774fb500aa2dbca34e6e1f5d9a14861
-size 5750506528
+oid sha256:9ba104c7049f27ecc91a76531c1a3dc782ee591101f9aab363a094fdc809e7be
+size 5750506496

Luau-Devstral-24B-Instruct-v0.1-IQ2_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c28fb6f3fa47c37ff0dba83f36614319314b574abf4b3854014e48ac6b2681a5
-size 8114062368
+oid sha256:f9f7fac543dd425a96f06488f4a7b511208ef201ac6e500af718dd6628074008
+size 8114062336

Luau-Devstral-24B-Instruct-v0.1-IQ3_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1dd1208ba71cf5a39c9d8655931db204053474af35f3dac094a02efabc5cdc13
-size 10650960928
+oid sha256:6e5a09df7f661b58b9ee93d9dbbd377c184ee208d99224b0a0442affa6790da2
+size 10650960896

Luau-Devstral-24B-Instruct-v0.1-IQ4_xs.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e671eb18e1592333f1fac3703645eb91445f098d05ebd410a200506b44919d48
+size 12758926336

Luau-Devstral-24B-Instruct-v0.1-Q3_K_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:551c49981550ec9cb11a6d969e447a1c6d17ed6f7d75f2a92aed4cb5e9186341
-size 11474093088
+oid sha256:9b5a44e23750727217bd6ce982c67374b75af5720acb892b108bc5883146c987
+size 11474093056

Luau-Devstral-24B-Instruct-v0.1-Q4_K_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c7a4eca76a1661a48a1028c1c8db55bffe140bc8cf2deb0a02fa31e29dcf30d5
-size 14333920288
+oid sha256:f430a622277f25ecb40a3d144e5aec2769bebdc831bd46a9cfcae7c217424992
+size 14333920256

Luau-Devstral-24B-Instruct-v0.1-Q5_K_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d2f6a684f607c4c572b35ec634805a7af62d68aa421f7125aae2aec9b52bde8
-size 16763995168
+oid sha256:113f1f3b6609bd3ce2acf93255740dede0a194cb6e0d3dc2fe18d25442fd8e96
+size 16763995136

Luau-Devstral-24B-Instruct-v0.1-Q6_K.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6a9ad3642573cc6057f791fefe29e9c51874f9a746785abc56702967edf394d0
-size 19345949728
+oid sha256:5e951fbff41020f3fd150648d676e9074e022190904ff1a1d93414a38793eb22
+size 19345949696

Luau-Devstral-24B-Instruct-v0.1-Q8_0.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:820ffe215ebdfeb6ad9e3057289b6c5b037073df1fbfe0598c9a1043c459d3c8
-size 25054790688
+oid sha256:a390a16f77f62e33c70b654f92ffca90d7e0c763cda9a1aad00f96c1f8ab03d2
+size 25054790656

README.md

@@ -11,7 +11,7 @@ language:
 - pt
 - it
 base_model:
-- unsloth/Devstral-Small-2507-unsloth-bnb-4bit
+- unsloth/Devstral-Small-2507
 tags:
 - roblox
 - luau
@@ -35,7 +35,7 @@ Devstral Small 2507 is a powerful choice for local inference, achieving SOTA ope
 - **Developed by:** Zack Williams ([boatbomber](https://huggingface.co/boatbomber))
 - **Funded by:** [Torpedo Software LLC](https://huggingface.co/TorpedoSoftware)
 - **License:** [Apache 2.0](https://www.tldrlegal.com/license/apache-license-2-0-apache-2-0)
-- **Finetuned from model:** [unsloth/Devstral-Small-2507-unsloth-bnb-4bit](https://huggingface.co/unsloth/Devstral-Small-2507-unsloth-bnb-4bit)
+- **Finetuned from model:** [unsloth/Devstral-Small-2507](https://huggingface.co/unsloth/Devstral-Small-2507)
 
 ### Model Sources
 
@@ -48,22 +48,31 @@ Devstral Small 2507 is a powerful choice for local inference, achieving SOTA ope
 ### Training Data
 
 1. https://huggingface.co/datasets/TorpedoSoftware/the-luau-stack
+    
+    Format:
+    ```md
+    Repository: {repo_name}
+    Repository Description: {repo_description}
+
+    File Path: `{file_path}`
+    File Content:
+    ```Lua
+    {file_content}
+    ```\
+    ```
 2. https://huggingface.co/datasets/TorpedoSoftware/roblox-info-dump
+    
+    Format:
+    ```md
+    Roblox Creator Docs: {url}
+    ```md
+    {content}
+    ```\
+    ```
 
-#### Preprocessing
+### Training Process
 
-Each datapoint from the training data was formatted as follows in order to provide the model with relevant context:
-
-```md
-Repository: {repo_name}
-Repository Description: {repo_description}
-
-File Path: `{file_path}`
-File Content:
-```Lua
-{file_content}
-```\
-```
+Trained a LoRA adapter (r=64) at full precision on two epochs of the dataset for a total of 54,630 steps and 43.40 E FLOPs. Then merged the final adapter checkpoint into a BF16 model. 
 
 ### Training Loss Curve
 
@@ -71,14 +80,12 @@ File Content:
 
 ### Imatrix Calibration
 
-The imatrix for the GGUF quantizations was computed using 33.5MB of text containing a combination of [wiki.train.raw](https://huggingface.co/datasets/ikawrakow/validation-datasets-for-llama.cpp/blob/main/wiki.train.raw.gz) and content from [the-luau-stack](https://huggingface.co/datasets/TorpedoSoftware/the-luau-stack) & [roblox-info-dump](https://huggingface.co/datasets/TorpedoSoftware/roblox-info-dump). This created an imatrix that is well suited to the specialized tasks this model is designed for while still maintaining broader intelligence as well. While we do provide several quantizations already, the `imatrix.gguf` is included in this repository should you want to create other quants yourself.
+The imatrix for the GGUF quantizations was computed using 5.73MB of text containing a combination of [technical.txt](https://huggingface.co/datasets/froggeric/imatrix/blob/main/technical.txt), [groups_merged.txt](huggingface.co/datasets/froggeric/imatrix/blob/main/groups_merged.txt), and content from [the-luau-stack](https://huggingface.co/datasets/TorpedoSoftware/the-luau-stack) & [roblox-info-dump](https://huggingface.co/datasets/TorpedoSoftware/roblox-info-dump). This created an imatrix that is well suited to the specialized tasks this model is designed for while still maintaining broader intelligence as well. While we do provide several quantizations already, the `imatrix.gguf` is included in this repository should you want to create other quants yourself.
 
 ## Environmental Impact
 
 Carbon emissions estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
 
-- **Hardware Type:** RTX 3090
+- **Hardware Type:** A100 80GB PCIe
 - **Hours used:** 60
-- **Cloud Provider:** My gaming PC
-- **Compute Region:** Bay Area
-- **Carbon Emitted:** 4.73 kg CO2eq (equivalent to 11.8 miles driven by an average ICE car)
+- **Carbon Emitted:** ~4.5 kg CO2eq (equivalent to ~10.1 miles driven by an average ICE car)

config.json

@@ -22,9 +22,9 @@
   "sliding_window": null,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.55.0",
+  "transformers_version": "4.55.3",
   "unsloth_fixed": true,
-  "unsloth_version": "2025.8.4",
+  "unsloth_version": "2025.8.9",
   "use_cache": true,
   "vocab_size": 131072
 }

generation_config.json

@@ -4,5 +4,5 @@
   "eos_token_id": 2,
   "max_length": 131072,
   "pad_token_id": 11,
-  "transformers_version": "4.55.0"
+  "transformers_version": "4.55.3"
 }

imatrix.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cf8786ed24584323884421b5dc7380c461dd4951af40299344c16dd9353a3da3
-size 10037312
+oid sha256:43e84a154059602f576de657d723d7f35aeb929660ee8f78238388e5d64e3f84
+size 10037344

model-00001-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68734b2d799fddfd97cd85ef845090ffee69471acb0ae07056d7c5e64fe60fc5
+oid sha256:24e3d5e7f129926fbbc91099bc357ed8dad2d535c00935b747738b97b5bc602a
 size 4781571736

model-00002-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c941455647f3f88ff79edc9a0c636d1c7f983e1af51eaaf6ec9c2bd8238bd4c
+oid sha256:7659bd8dc9c015c6ed3cb6acfdbabe3195eb33aac6475ddc1a2d080b715eea78
 size 4781592784

model-00003-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cdf6c38f239f14d5accec4f2e3528d6c6dd50f9b60d0b1126e11c6b8d43502c7
+oid sha256:da3d9b538cd7c9a689ca42819a935f2654f4e69dbbdb9ab27e8f264e1a504b81
 size 4781592800

model-00004-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:95b81100943834323de390b4ce8c04c817cb7f77e3e9872f38ecabd45028afde
+oid sha256:71cd682ad99eb857404b24616bbaa9b9fc9741a62d1c4421ba6bcd5a9d55944b
 size 4886471600

model-00005-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:38b4fbc7ddb9ae5959c03d3de681a6f88ce986a81b7d4a148193a20c758ab97d
+oid sha256:c364dc773505c6eaa70ecde6f2c0a4fc8cade2921d978d8f1a3190a7121e3b45
 size 4781592824

model-00006-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dd2e644e62dd3eb0c41795e030402107732ce3924d80b7ba282d942698f44d63
+oid sha256:dda03a2c951066cc3b5d84bf939264fa94557ffb1a37da42cadeac3cc95f0eb0
 size 4781592816

model-00007-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:92fb788f5ded06652f0c2af61bd797e618b7f7af8e1a6affe80fdf4258977486
+oid sha256:dc5527a3f6c7a736129e275e166ea054b327ac74fb16e891feabfd2088c1aea0
 size 4886471600

model-00008-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d4c2b3daaddf1617c3fdda3427944ec3088350f6e2de61f3224a84132ce6645
+oid sha256:e24bbb48a307739507bef06531ede237884fbb4c0f9e3d5c33b72b9ba2edc22f
 size 4781592824

model-00009-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:827e4a75e28e786c2fd54fe18461a7dcb86925b1358dbf540d98c3a412105ef7
+oid sha256:735e74b7c38a12cd2d607514d50d5c8591c22fc88baa153396ba9c393562c060
 size 4781592816

model-00010-of-00010.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d436770ac3f8e5c2c2d051a22345498b95584b5837d65051d7ab07834f91d46
+oid sha256:7a57790fb693007a0af0af4351dbc0a6367f25789d9422874f8d50cae1a487b0
 size 3900777072

model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
   "metadata": {
-    "total_parameters": 23757214720,
+    "total_parameters": 23572403200,
     "total_size": 47144806400
   },
   "weight_map": {

tokenizer_config.json

@@ -9013,7 +9013,7 @@
   "legacy": true,
   "model_max_length": 131072,
   "pad_token": "<pad>",
-  "padding_side": "left",
+  "padding_side": "right",
   "tokenizer_class": "LlamaTokenizerFast",
   "unk_token": "<unk>",
   "use_default_system_prompt": false