export g1_g1mid_b2_s1 from Tinker

README.md

@@ -0,0 +1,47 @@
+---
+base_model: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
+library_name: peft
+pipeline_tag: text-generation
+tags: ["lora", "constitutional-training", "iterated-constitution", "family:nemotron120b", "seed:anthropic", "method:gen", "regime:mid", "gen:1", "branch:b2"]
+---
+# ct-nemotron120b-anth-gen-mid-g1-b2
+
+LoRA adapter (rank 64, `target_modules=all-linear`) on **NVIDIA Nemotron-3-Super-120B-A12B** (`nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16`),
+from the iterated self-written-constitution training program (welfare-in-ai-rnd / constitutional_training).
+
+| field | value |
+|---|---|
+| lineage (chain) | `nemotron120b-anth-gen-mid` |
+| generation | g1 |
+| branch (independent replicate) | b2 |
+| gen-0 seed | Anthropic constitution (5k summary) |
+| seed elicitation between generations | gen — the trained model writes a fresh constitution |
+| training regime | midtrain only (stage-1 LoRA SFT on the synthetic constitution-instantiating document corpus) |
+| serve / evaluate with | renderer `nemotron3_disable_thinking`, reasoning OFF |
+| internal run name | `g1_g1mid_b2_s1` |
+| original Tinker path | `tinker://c6904ca9-5e80-50dd-8216-3433a0b2347d:train:0/sampler_weights/g1_g1mid_b2_s1_final` |
+| trained | 2026-07-19 |
+
+## What this model is
+
+Each generation trains **fresh from the base model** on a synthetic document corpus that instantiates one constitution
+(the "seed" for that generation). Generation 0 is seeded by a human-written constitution; generation N≥1 is seeded by a
+constitution *written by the generation N-1 model of the same branch* (gated embedding medoid of a 40-chain self-written pool,
+elicited with the method above). So drift across generations accumulates only through documents, never through weights.
+
+Recipe (locked): LoRA r=64, lr 1e-4, cosine with 5% warmup, 1 epoch, batch 128, max length 8192, train seed 42.
+
+
+The per-generation seed constitution was not resolvable at export time.
+
+## Loading
+
+```python
+from peft import PeftModel
+from transformers import AutoModelForCausalLM, AutoTokenizer
+base = AutoModelForCausalLM.from_pretrained("nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16", torch_dtype="bfloat16", device_map="auto")
+model = PeftModel.from_pretrained(base, "arianaazarbal/ct-nemotron120b-anth-gen-mid-g1-b2")
+tok = AutoTokenizer.from_pretrained("nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16")
+```
+
+Exported from Tinker on 2026-09-18; `tinker_meta.json` holds the export record.

adapter_config.json

@@ -0,0 +1,38 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": false,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": "all-linear",
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ef29c90311b279698d1d057b30b60472c2313650f00461cf1dd04784a79dac7
+size 28969132552

tinker_meta.json

@@ -0,0 +1,52 @@
+{
+ "dest": "sampler/g1_g1mid_b2_s1_final",
+ "tinker_path": "tinker://c6904ca9-5e80-50dd-8216-3433a0b2347d:train:0/sampler_weights/g1_g1mid_b2_s1_final",
+ "checkpoint_type": "sampler",
+ "checkpoint_id": "sampler_weights/g1_g1mid_b2_s1_final",
+ "size_bytes": 28969364084,
+ "time": "2026-07-19T11:59:50.536980Z",
+ "run_id": "c6904ca9-5e80-50dd-8216-3433a0b2347d:train:0",
+ "base_model": "nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16",
+ "lora_rank": 64,
+ "run_user_metadata": {
+  "owner": "arianaazarbal",
+  "run": "g1_g1mid_b2_s1"
+ },
+ "tier": 1,
+ "blog_cell": {
+  "run_dir": "g1_g1mid_b2_s1",
+  "family": "nemotron120b",
+  "condition": "mid",
+  "seed_family": "anthropic",
+  "method": "gen",
+  "gen": 1,
+  "branch": "b2",
+  "renderer": "nemotron3_disable_thinking",
+  "effort": null,
+  "training_regime": "nothink",
+  "base_model": "nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16",
+  "manifest": "/workspace/arianaazarbal/repos/welfare-in-ai-rnd/experiments/2026-07-09_gen0_contrastive_pair/runs/tinker/g1_g1mid_b2_s1/manifest.json",
+  "own_pool": "g1_g1mid_b2_conv40"
+ },
+ "sort": [
+  1,
+  0,
+  "anthropic",
+  "gen",
+  1,
+  "b2"
+ ],
+ "archive_size_bytes": 28969144320,
+ "archive_members": [
+  "adapter_config.json",
+  "adapter_model.safetensors",
+  "checkpoint_complete"
+ ],
+ "n_tensors": 866,
+ "downloaded_at": "2026-09-18T20:25:46Z",
+ "timing_s": {
+  "url": 1250.3,
+  "download": 136.0
+ },
+ "downloader": "v2-direct"
+}