Text Generation
MLX
Safetensors
English
small-language-model
from-scratch
conversational
apple-silicon
research
Instructions to use OpenSML/Solstice-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use OpenSML/Solstice-1 with MLX:
# Make sure mlx-lm is installed # pip install --upgrade mlx-lm # Generate text with mlx-lm from mlx_lm import load, generate model, tokenizer = load("OpenSML/Solstice-1") prompt = "Write a story about Einstein" messages = [{"role": "user", "content": prompt}] prompt = tokenizer.apply_chat_template( messages, add_generation_prompt=True ) text = generate(model, tokenizer, prompt=prompt, verbose=True) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- MLX LM
How to use OpenSML/Solstice-1 with MLX LM:
Generate or start a chat session
# Install MLX LM uv tool install mlx-lm # Interactive chat REPL mlx_lm.chat --model "OpenSML/Solstice-1"
Run an OpenAI-compatible server
# Install MLX LM uv tool install mlx-lm # Start the server mlx_lm.server --model "OpenSML/Solstice-1" # Calling the OpenAI-compatible server with curl curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "OpenSML/Solstice-1", "messages": [ {"role": "user", "content": "Hello"} ] }' - Atomic Chat
File size: 11,194 Bytes
7bcd331 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 | {
"allowed_test_manifests": [
"2118c7446c412a4ee7a4b7a356a211ad91574983e213da45980707b5fa1ca124",
"f9273450799963b2fd8e61a5e9cb6892be7ba4a2308849623e288302a687a1f7",
"dcab4ad8e4484ad1cc4b32ffbc6352b6a6ff9ea0a4ddccdb8d367d23db098a3f"
],
"selection": "Highest predeclared development key; public and final-test results cannot change this choice",
"test_data_sha256": "abcdf03853c4f00fff1e92cdf5f98d8ee3bea42526c18f15585650b368ddd2f3",
"winner": {
"attempt": 4,
"checkpoint": "${WORKSPACE}/pretraining/runs/deep44_chat_repair_04_20261109/step_0000128_5937a6def881",
"checkpoint_manifest": "2118c7446c412a4ee7a4b7a356a211ad91574983e213da45980707b5fa1ca124",
"legacy_dev": {
"assistant_nll": 0.9992501850956614,
"components": {
"constraints": 0.35257945707417093,
"dialogue": 0.6570782772469101,
"general": 1.7640471532940865,
"grounded": 0.41334675560938194,
"knowledge": 1.3135829586535692,
"rewrite": 1.294418379664421,
"skills": 0.592298005183693,
"social": 1.6066504940390587
},
"empty_rate": 0.0,
"generated_counts": {
"constraints": 64,
"dialogue": 64,
"grounded": 64,
"knowledge": 64,
"skills": 64
},
"generated_macro_pass": 0.625,
"generated_pass_counts": {
"constraints": 56,
"dialogue": 59,
"grounded": 42,
"knowledge": 13,
"skills": 30
},
"generated_rates": {
"constraints": 0.875,
"dialogue": 0.921875,
"grounded": 0.65625,
"knowledge": 0.203125,
"skills": 0.46875
},
"natural_stop": 0.9861111111111112,
"precision": "fp32-chat-repair",
"qualitative_count": 40,
"repetition_rate": 0.022222222222222223,
"replacement_character_rate": 0.0,
"retention": {
"sources": {
"dclm": 2.9645858108997345,
"web": 2.8973595052957535,
"wiki": 2.422679301351309
},
"val_loss": 2.8633367460221053,
"validation_fingerprint": "2e9444ddc7de3cf4685911d967c1ad605d9b8cfd9244f7039621e828bf088ffa"
},
"scope": "Checked grounded answers, deterministic tasks and authored conversation context. Public general/dialogue/social outputs require semantic review. No code, SQL, JSON, or public benchmark scores select checkpoints.",
"split": "dev",
"step": 128,
"stop_reason_counts": {
"token_limit": 5,
"turn_end": 355
},
"turn_end_rate": 0.9861111111111112
},
"new_dev": {
"correct": 179,
"counts": {
"abstention": 57,
"constraints": 6,
"dialogue": 58,
"grounded": 46,
"reasoning": 12
},
"empty": 0,
"false_abstention_signals": 13,
"identity": {
"data": "a9c20f71859a9269b4875ef301e45938bb8d36ee62f3d135dcddc1ac6f14f066",
"evaluator": "40dbfada2856c91b0a7ffe5df7b9584175dbd479a1c6b581d1935d38dd6228aa",
"parent": null
},
"missing_fact_abstention_signals": 57,
"natural_stop": 0.9966666666666667,
"repetition_rate": 0.0033333333333333335,
"scoring": "Conservative whole-answer equivalence for factual tasks; exact format for transformation tasks. Strict format compliance is reported separately. Whole-passage copying and contradictory extra claims are not accepted.",
"strict_correct": 179,
"strict_counts": {
"abstention": 57,
"constraints": 6,
"dialogue": 58,
"grounded": 46,
"reasoning": 12
},
"total": 300,
"totals": {
"abstention": 60,
"constraints": 60,
"dialogue": 60,
"grounded": 60,
"reasoning": 60
}
},
"parent": "attempt5",
"plan": {
"adaptation": {
"adapter_compute": "float32",
"alpha": 64,
"base_compute": "bfloat16",
"dropout": 0.0,
"expected_trainable_parameters": 3785856,
"layers": "all",
"method": "lora",
"rank": 32,
"targets": [
"attn.qkv_proj",
"attn.proj"
],
"trainable_token_ids": [
2,
3
]
},
"answer_max_tokens": {
"constraints": 320,
"dialogue": 256,
"general": 256,
"grounded": 160,
"knowledge": 80,
"rewrite": 320,
"skills": 160,
"social": 128
},
"authored_holdouts": {
"constraints": 128,
"dialogue": 128,
"skills": 192
},
"authored_quotas": {
"constraints": 4200,
"dialogue": 6800,
"skills": 12000
},
"baseline_cache": "${WORKSPACE}/pretraining/prepared/chat_adapter_search_20261108/baseline_cache",
"batches": [
8
],
"betas": [
0.9,
0.95
],
"checked_components": [
"knowledge",
"grounded",
"skills",
"constraints",
"dialogue"
],
"comparison_checkpoint": "${WORKSPACE}/reports/benchmarks/deep44-reliable-chat-167-fp32-20261106/choices/checkpoint",
"component_quotas": {
"constraints": 800,
"dialogue": 2000,
"general": 2400,
"grounded": 2000,
"skills": 800
},
"components": [
"general",
"knowledge",
"grounded",
"skills",
"constraints",
"dialogue",
"social",
"rewrite"
],
"epochs": 1,
"eps": 1e-08,
"eval_every": 64,
"eval_examples_per_component": 64,
"execution": "single-mac",
"final_lr": 2.0000000000000003e-06,
"format": "opensml-verified-chat-v2",
"fresh_mixture": {
"abstention": 0.1,
"dialogue": 0.15,
"grounded": 0.15,
"reasoning": 0.6
},
"fresh_weight": 0.4,
"generated_eval_per_component": 64,
"grad_accum": 8,
"grad_clip": 1.0,
"initial_checkpoint": "${WORKSPACE}/pretraining/runs/deep44_parent2000_generalize_05_replica_20261109/step_0000256_6074b98fe66b",
"keep_checkpoints": 6,
"kl_weight": 0.25,
"log_every": 10,
"max_cumulative_retention_loss_increase": 0.03,
"max_messages": 8,
"max_repetition_rate": 0.025,
"max_retention_loss_increase": 0.003,
"max_sequence_length": 2048,
"max_steps": 256,
"method": "Short checked reasoning",
"microbatch_rows": 4,
"min_domain_pass_cases": {},
"min_natural_stop": 0.9861111111111112,
"min_task_gain_cases": 4,
"minimum_source_tokens": 42000000000,
"original_pretraining_retention_loss": 2.8353193145245315,
"output": "${WORKSPACE}/pretraining/runs/deep44_chat_repair_04_20261109",
"peak_lr": 2e-05,
"prepared": "${WORKSPACE}/pretraining/runs/deep44_chat_repair_04_20261109/input/corpus",
"public_holdouts": {
"constraints_smol": 32,
"dialogue_smol": 64,
"general_smol": 64,
"grounded_no": 32,
"grounded_yes": 96,
"knowledge": 128,
"rewrite_smol": 32,
"social": 32
},
"public_quotas": {
"constraints_smol": 2000,
"dialogue_smol": 1200,
"general_smol": 4800,
"grounded_no": 2750,
"grounded_yes": 8250,
"knowledge": 4000,
"rewrite_smol": 1000,
"social": 1000
},
"qualitative_eval_per_component": 8,
"regression_patience": 2,
"repair_parent": "attempt5",
"replay_length": 512,
"replay_mixture": {
"constraints": 0.125,
"dialogue": 0.125,
"general": 0.1875,
"grounded": 0.125,
"knowledge": 0.1875,
"rewrite": 0.0625,
"skills": 0.125,
"social": 0.0625
},
"replay_objective_weight": 0.2,
"replay_policy": "prose-only",
"replay_rows_per_update": 16,
"retention": "${WORKSPACE}/pretraining/data/deep44_pretraining_validation_20261102",
"rows_per_group": 16,
"seed": 2026110915,
"selection_reference": "New 300-case development test, with historical 320-case and prose retention guards. Final 1000-case test is unopened until a single finalist is frozen.",
"sft_objective_weight": 0.8,
"sft_rows_per_update": 48,
"sft_stage": 2,
"source_checkpoint": "${WORKSPACE}/pretraining/runs/deep44_cooldown_2b_20261107/step_0320445_08b689895ed2",
"source_run": "${WORKSPACE}/pretraining/runs/deep44_cooldown_2b_20261107",
"sources": [
{
"file": "train.jsonl.gz",
"label": "helpsteer2",
"license": "cc-by-4.0",
"repo": "nvidia/HelpSteer2",
"revision": "990b2711a36180dd19d9c94b8627844866f8982a"
},
{
"file": "data/train-00000-of-00001-b42a775f407cee45.parquet",
"label": "oasst1",
"license": "apache-2.0",
"repo": "OpenAssistant/oasst1",
"revision": "fdf72ae0827c1cda404aff25b6603abec9e3399b"
},
{
"files": [
"data/smol-magpie-ultra/train-00000-of-00006.parquet",
"data/smol-magpie-ultra/train-00001-of-00006.parquet",
"data/smol-magpie-ultra/train-00002-of-00006.parquet",
"data/smol-magpie-ultra/train-00003-of-00006.parquet",
"data/smol-magpie-ultra/train-00004-of-00006.parquet",
"data/smol-magpie-ultra/train-00005-of-00006.parquet"
],
"label": "smol_magpie_ultra",
"license": "apache-2.0",
"repo": "HuggingFaceTB/smoltalk",
"revision": "5feaf2fd3ffca7c237fc38d1861bc30365d48ffa"
},
{
"label": "squad",
"license": "cc-by-sa-4.0",
"parquet_urls": [
"https://huggingface.co/datasets/rajpurkar/squad_v2/resolve/3ffb306f725f7d2ce8394bc1873b24868140c412/squad_v2/train-00000-of-00001.parquet"
],
"repo": "rajpurkar/squad_v2",
"revision": "3ffb306f725f7d2ce8394bc1873b24868140c412",
"split": "train"
},
{
"label": "squad_validation",
"license": "cc-by-sa-4.0",
"parquet_urls": [
"https://huggingface.co/datasets/rajpurkar/squad_v2/resolve/3ffb306f725f7d2ce8394bc1873b24868140c412/squad_v2/validation-00000-of-00001.parquet"
],
"repo": "rajpurkar/squad_v2",
"revision": "3ffb306f725f7d2ce8394bc1873b24868140c412",
"split": "validation"
}
],
"strategy": "Independent small adapter updates from parent2000 or preserved attempt5; verified text tasks and 60% replay objective with frozen-source probability preservation.",
"tokenizer": "${WORKSPACE}/reference/opensml-150m/tokenizer/bytebpe32k_v1",
"warmup_fraction": 0.1,
"weight_decay": 0.0
},
"quality": {
"consecutive_severe": 0,
"eligible": true,
"key": [
1,
179,
200,
0.9966666666666667,
-0.0033333333333333335,
-0.9992501850956614
],
"reasons": []
},
"status": "complete",
"step": 128
}
}
|