Spaces:
Sleeping
Sleeping
File size: 38,918 Bytes
e0670a4 8ff0c80 e0670a4 4b8c9b5 e0670a4 4b8c9b5 e0670a4 4b8c9b5 e0670a4 4b8c9b5 e0670a4 4b8c9b5 e0670a4 4b8c9b5 e0670a4 66bd94a e0670a4 66bd94a e0670a4 66bd94a e0670a4 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 | """AdaptiveRAG β under-the-hood pipeline visualizer.
Run: streamlit run app.py
"""
from __future__ import annotations
import json
import logging
import os
import tempfile
import time
from pathlib import Path
os.environ.setdefault("ANONYMIZED_TELEMETRY", "False")
# suppress harmless noise from Streamlit's torch inspector + ChromaDB posthog client
logging.getLogger("streamlit.watcher.local_sources_watcher").setLevel(logging.ERROR)
logging.getLogger("chromadb.telemetry.product.posthog").setLevel(logging.ERROR)
import numpy as np
import pandas as pd
import streamlit as st
from agent.critic import critique, refine_query
from agent.planner import plan
from agent.router import route
from agent.tools import image_retrieve_and_reason
from config import AGENT_CONFIG, EMBEDDING_CONFIG, HOSTED, LLM_CONFIG, PATHS, RETRIEVAL_CONFIG
from ingestion.embedder import embed_query
from ingestion.indexer import fetch_embeddings
from llm.client_factory import get_llm
from retrieval.dense import Hit, dense_search
from retrieval.hybrid import reciprocal_rank_fusion
from retrieval.reranker import rerank
from retrieval.sparse import sparse_search
st.set_page_config(page_title="AdaptiveRAG β Underhood", page_icon="π¬", layout="wide")
# ββ LLM backend check β shown before anything else ββββββββββββββββββ
if not os.environ.get("GROQ_API_KEY"):
# Running without Groq β check if Ollama is reachable locally
try:
import requests as _req
_req.get("http://localhost:11434/api/tags", timeout=2).raise_for_status()
_ollama_ok = True
except Exception:
_ollama_ok = False
if not _ollama_ok:
st.error(
"**No LLM backend found.**\n\n"
"- **Running on Hugging Face?** Add your `GROQ_API_KEY` secret in "
"Space Settings β Variables and secrets. Get a free key at "
"[console.groq.com](https://console.groq.com).\n"
"- **Running locally?** Start Ollama: `ollama serve`"
)
st.stop()
# βββββββββββββββββββββββββββββ styling ββββββββββββββββββββββββββββββ
st.markdown(
"""
<style>
.phase-card {
border-left: 4px solid var(--accent, #4f8cff);
padding: .6rem 1rem;
margin: .25rem 0 .5rem 0;
background: rgba(79,140,255,0.06);
border-radius: 6px;
}
.phase-num { color: #4f8cff; font-weight: 700; margin-right: .4rem; }
.pill { display: inline-block; padding: .15rem .55rem; border-radius: 999px;
font-size: .78rem; font-weight: 600; margin-right: .4rem; }
.pill-blue { background: #1e3a5f; color: #9ec5ff; }
.pill-green { background: #1e4f30; color: #a3e6b5; }
.pill-purple { background: #3d2a5e; color: #c8a8f5; }
.pill-amber { background: #5e3f0e; color: #f3c97a; }
.pill-red { background: #5a1f1f; color: #f3a3a3; }
.pill-grey { background: #2c2c33; color: #b8b8c0; }
.chunk-card {
background: rgba(255,255,255,0.03);
border: 1px solid rgba(255,255,255,0.08);
border-radius: 6px; padding: .55rem .7rem; margin-bottom: .4rem;
font-size: .82rem;
}
.chunk-meta { color: #9aa3b2; font-size: .73rem; margin-bottom: .25rem; }
.mini-vec {
font-family: 'SF Mono', Menlo, monospace; font-size: .68rem;
color: #8b949e; word-break: break-all;
}
</style>
""",
unsafe_allow_html=True,
)
# βββββββββββββββββββββββββββββ helpers ββββββββββββββββββββββββββββββ
@st.cache_resource
def _llm():
return get_llm()
def _load_manifest() -> dict:
p = PATHS["manifest_path"]
return json.loads(p.read_text()) if p.exists() else {}
def phase_header(num: int, title: str, subtitle: str = "") -> None:
st.markdown(
f"<div class='phase-card'><span class='phase-num'>STEP {num}</span>"
f"<b>{title}</b><br><span style='color:#9aa3b2;font-size:.85rem;'>{subtitle}</span>"
f"</div>",
unsafe_allow_html=True,
)
def hits_to_df(hits: list[Hit], score_label: str = "score") -> pd.DataFrame:
rows = []
for h in hits:
title = h.metadata.get("title") or h.metadata.get("source_path", "?")
short = title.split(" (")[0]
if len(short) > 38:
short = short[:35] + "β¦"
label = f"{short} Β· p{h.metadata.get('page_start')} Β· {h.chunk_id.split('::')[-1]}"
rows.append({"chunk": label, score_label: float(h.score), "chunk_id": h.chunk_id})
return pd.DataFrame(rows)
def render_hits(hits: list[Hit], badge_class: str, label: str, max_chars: int = 220) -> None:
if not hits:
st.caption(f"_(no {label.lower()} hits)_")
return
for i, h in enumerate(hits, start=1):
meta = h.metadata
snippet = h.text[:max_chars].replace("\n", " ")
if len(h.text) > max_chars:
snippet += "β¦"
st.markdown(
f"<div class='chunk-card'>"
f"<div class='chunk-meta'>"
f"<span class='pill {badge_class}'>{label} #{i}</span>"
f"score <b>{h.score:.3f}</b> Β· "
f"{meta.get('title','?')} Β· p.{meta.get('page_start')}β{meta.get('page_end')} Β· "
f"<code>{h.chunk_id}</code>"
f"</div>{snippet}</div>",
unsafe_allow_html=True,
)
def pca_2d(matrix: np.ndarray) -> np.ndarray:
centered = matrix - matrix.mean(axis=0, keepdims=True)
_, _, vt = np.linalg.svd(centered, full_matrices=False)
return centered @ vt[:2].T
def vector_space_plot(query_vec: list[float], fused_hits: list[Hit],
dense_ids: set[str], sparse_ids: set[str],
kept_ids: set[str]) -> None:
if not fused_hits:
st.caption("_(nothing to plot)_")
return
embs = fetch_embeddings([h.chunk_id for h in fused_hits])
rows = []
vecs = [np.array(query_vec, dtype=np.float32)]
for h in fused_hits:
v = embs.get(h.chunk_id)
if v is None:
continue
vecs.append(np.array(v, dtype=np.float32))
in_d, in_s = h.chunk_id in dense_ids, h.chunk_id in sparse_ids
in_keep = h.chunk_id in kept_ids
if in_d and in_s:
color = "fused (both)"
elif in_d:
color = "dense only"
elif in_s:
color = "sparse only"
else:
color = "other"
title = (h.metadata.get("title") or "?").split(" (")[0][:40]
label = f"{title} Β· p{h.metadata.get('page_start')}"
rows.append({"label": label, "color": color, "size": 90 if in_keep else 50})
if len(vecs) < 3:
st.caption("_(need at least 2 hits for a 2D projection)_")
return
proj = pca_2d(np.vstack(vecs))
df = pd.DataFrame(
[{"x": proj[0, 0], "y": proj[0, 1], "label": "π your question",
"color": "QUERY", "size": 220}]
+ [{"x": proj[i + 1, 0], "y": proj[i + 1, 1], **rows[i]}
for i in range(len(rows))]
)
st.scatter_chart(
df, x="x", y="y", color="color", size="size",
height=380, use_container_width=True,
)
st.caption(
"PCA projection of the query embedding + fused hit embeddings. "
"Larger points survived cross-encoder reranking."
)
def render_embedding_card(query: str, qv: list[float], dt: float) -> None:
arr = np.array(qv, dtype=np.float32)
cols = st.columns([1, 1, 1, 3])
cols[0].metric("Model", EMBEDDING_CONFIG["model"].split("/")[-1])
cols[1].metric("Dimensions", len(qv))
cols[2].metric("L2 norm", f"{float(np.linalg.norm(arr)):.3f}")
cols[3].metric("Embed time", f"{dt*1000:.0f} ms")
st.caption(f"Question ({len(query)} chars, ~{len(query.split())} words):")
st.code(query, language="text")
st.caption("First 32 dimensions of the embedding vector:")
st.bar_chart(pd.DataFrame({"value": arr[:32]}), height=140, use_container_width=True)
preview = ", ".join(f"{x:+.3f}" for x in arr[:8]) + ", β¦"
st.markdown(f"<span class='mini-vec'>vector[0:8] = [{preview}]</span>",
unsafe_allow_html=True)
# βββββββββββββββββββββββββββββ pipeline view ββββββββββββββββββββββββββββββ
def _render_healing_trace(healing_trace: list[dict], health_score: float) -> None:
"""Render the healing trace panel below the answer."""
# Health score metric with colour
if health_score >= 80:
color, label = "#2ecc71", "Healthy"
elif health_score >= 60:
color, label = "#f39c12", "Fair"
else:
color, label = "#e74c3c", "Needs healing"
st.markdown(
f"<div style='display:flex;align-items:center;gap:1rem;margin:.5rem 0;'>"
f"<span style='font-size:1.1rem;font-weight:700;color:{color};'>"
f"βοΈ Health score: {health_score:.0f} / 100</span>"
f"<span class='pill' style='background:{color}22;color:{color};border:1px solid {color}44;'>"
f"{label}</span></div>",
unsafe_allow_html=True,
)
if not healing_trace:
st.success("β
Answer passed all checks on first attempt β no healing needed.")
return
for attempt in healing_trace:
num = attempt.get("attempt", "?")
healthy = attempt.get("healthy", False)
icon = "β
" if healthy else "π§"
issues = attempt.get("issues", [])
label_str = "Healthy" if healthy else f"Issues: {', '.join(issues)}"
with st.expander(f"{icon} Attempt {num} β {label_str}", expanded=not healthy):
if healthy:
st.success("All checks passed β answer accepted.")
continue
c1, c2, c3 = st.columns(3)
c1.metric("Hallucination", "β οΈ yes" if "hallucination" in issues else "β
no")
c2.metric("Low chunk quality", "β οΈ yes" if "low_chunk_quality" in issues else "β
no")
c3.metric("Knowledge gap", "β οΈ yes" if "knowledge_gap" in issues else "β
no")
for action in attempt.get("actions", []):
atype = action.get("type", "")
detail = action.get("detail", "")
icons = {
"hallucination_fix": "π",
"chunk_expansion": "π",
"web_search": "π",
"query_rewrite": "βοΈ",
"regenerate": "π",
}
prefix = icons.get(atype, "β’")
st.markdown(
f"<div class='chunk-card'><span class='pill pill-grey'>{atype}</span>"
f"{prefix} {detail}</div>",
unsafe_allow_html=True,
)
if atype == "hallucination_fix":
for s in action.get("flagged_sentences", []):
st.caption(f" β³ Flagged: \"{s[:120]}β¦\"")
def visual_pipeline(query: str, enable_healing: bool = True) -> None:
llm = _llm()
# ββ Step 1: embed the question ββββββββββββββββββββββββββββββββ
phase_header(1, "Question encoding",
"Convert text β 384-dim dense vector via sentence-transformers (MiniLM-L6).")
t0 = time.time()
qv = embed_query(query)
render_embedding_card(query, qv, time.time() - t0)
# ββ Step 2: Self-RAG router ββββββββββββββββββββββββββββββββ
phase_header(2, "Self-RAG router",
"Decide whether to RETRIEVE, ANSWER_DIRECTLY, or CLARIFY before touching the index.")
t0 = time.time()
decision = route(query, llm=llm)
dt = time.time() - t0
pill_map = {"RETRIEVE": "pill-blue", "ANSWER_DIRECTLY": "pill-green", "CLARIFY": "pill-amber"}
pill = pill_map.get(decision["action"], "pill-grey")
st.markdown(
f"<span class='pill {pill}'>{decision['action']}</span>"
f"<span style='color:#9aa3b2;'>{decision.get('reason','')}</span>"
f"<span style='float:right;color:#9aa3b2;font-size:.78rem;'>"
f"router latency: {dt*1000:.0f} ms</span>",
unsafe_allow_html=True,
)
if decision["action"] == "ANSWER_DIRECTLY":
st.markdown("### Direct answer (no retrieval)")
ans = llm.generate(prompt=query,
system="You are a helpful research assistant. Be concise.",
temperature=0.2)
st.markdown(ans)
return
if decision["action"] == "CLARIFY":
st.markdown("### Clarifying question")
ans = llm.generate(
prompt=("The user asked: " + query +
"\n\nIt is too ambiguous to answer well. Ask one short clarifying question."),
system="You are a helpful research assistant.",
temperature=0.2,
)
st.markdown(ans)
return
# ββ Iterations of plan β retrieve β answer β critique ββββββββ
accumulated: list[Hit] = []
current_query = query
for it in range(AGENT_CONFIG["max_iterations"]):
st.markdown(f"---\n## π Iteration {it + 1}")
if current_query != query:
st.info(f"Refined query β **{current_query}**")
# ββ Step 3: plan βββββββββββββββββββββββββββββββββββββ
phase_header(3, "Planner", "LLM decomposes the question into focused sub-queries.")
prior = ""
if accumulated:
titles = sorted({h.metadata.get("title", "?") for h in accumulated})
prior = "Already gathered passages from: " + ", ".join(titles)
t0 = time.time()
steps = plan(current_query, prior_summary=prior, llm=llm)
dt = time.time() - t0
st.caption(f"Generated {len(steps)} sub-quer{'y' if len(steps)==1 else 'ies'} in {dt*1000:.0f} ms")
for i, s in enumerate(steps, start=1):
st.markdown(
f"<div class='chunk-card'>"
f"<span class='pill pill-purple'>sub-query {i}</span>"
f"<b>{s['query']}</b>"
f"<div class='chunk-meta' style='margin-top:.3rem;'>"
f"rationale: {s.get('rationale','β')}</div></div>",
unsafe_allow_html=True,
)
# ββ Step 4: retrieval per sub-query ββββββββββββββββββ
phase_header(
4,
"Hybrid retrieval per sub-query",
f"Dense (Chroma cosine, k={RETRIEVAL_CONFIG['dense_k']}) β₯ "
f"Sparse (BM25, k={RETRIEVAL_CONFIG['sparse_k']}) β "
f"Reciprocal Rank Fusion β Cross-encoder rerank "
f"(BGE, top {RETRIEVAL_CONFIG['rerank_top_n']}).",
)
for si, step in enumerate(steps, start=1):
with st.expander(f"Sub-query {si}: {step['query']}", expanded=(si == 1)):
t0 = time.time()
dense_hits = dense_search(step["query"])
t_dense = time.time() - t0
t0 = time.time()
sparse_hits = sparse_search(step["query"])
t_sparse = time.time() - t0
t0 = time.time()
fused = reciprocal_rank_fusion([dense_hits, sparse_hits],
top_k=max(RETRIEVAL_CONFIG["dense_k"],
RETRIEVAL_CONFIG["sparse_k"]))
t_fuse = time.time() - t0
t0 = time.time()
reranked = rerank(step["query"], fused)
t_rerank = time.time() - t0
m1, m2, m3, m4 = st.columns(4)
m1.metric("Dense hits", len(dense_hits), f"{t_dense*1000:.0f} ms")
m2.metric("Sparse hits", len(sparse_hits), f"{t_sparse*1000:.0f} ms")
m3.metric("After RRF", len(fused), f"{t_fuse*1000:.0f} ms")
m4.metric("After rerank", len(reranked), f"{t_rerank*1000:.0f} ms")
tabs = st.tabs([
"π΅ Dense (vectors)",
"π’ Sparse (BM25)",
"π£ RRF fusion",
"π‘ Cross-encoder rerank",
"πΊοΈ Vector space",
])
with tabs[0]:
st.caption("Top-K nearest neighbors by cosine similarity.")
if dense_hits:
st.bar_chart(hits_to_df(dense_hits, "cosine_sim"),
x="chunk", y="cosine_sim",
height=260, use_container_width=True)
render_hits(dense_hits[:5], "pill-blue", "DENSE")
with tabs[1]:
st.caption("Top-K BM25 keyword matches (normalized).")
if sparse_hits:
st.bar_chart(hits_to_df(sparse_hits, "bm25_norm"),
x="chunk", y="bm25_norm",
height=260, use_container_width=True)
render_hits(sparse_hits[:5], "pill-green", "BM25")
with tabs[2]:
st.caption(
"Reciprocal Rank Fusion: score(d) = Ξ£ 1/(k + rank). "
"Combines dense + sparse rankings into one merged list."
)
if fused:
st.bar_chart(hits_to_df(fused[:12], "rrf_score"),
x="chunk", y="rrf_score",
height=280, use_container_width=True)
render_hits(fused[:5], "pill-purple", "FUSED")
with tabs[3]:
st.caption(
"Cross-encoder scores (query, chunk) jointly β much more "
"accurate than bi-encoder cosine, but slower β only run on "
"the fused candidate set."
)
if reranked:
st.bar_chart(hits_to_df(reranked, "ce_score"),
x="chunk", y="ce_score",
height=240, use_container_width=True)
render_hits(reranked, "pill-amber", "RERANKED")
with tabs[4]:
dense_ids = {h.chunk_id for h in dense_hits}
sparse_ids = {h.chunk_id for h in sparse_hits}
kept_ids = {h.chunk_id for h in reranked}
vector_space_plot(qv, fused[:20], dense_ids, sparse_ids, kept_ids)
accumulated.extend(reranked)
# ββ Step 5: answer βββββββββββββββββββββββββββββββββββββ
# Dedupe + cap to 8 passages for the final prompt
seen: set[str] = set()
unique: list[Hit] = []
for h in accumulated:
if h.chunk_id in seen:
continue
seen.add(h.chunk_id)
unique.append(h)
if len(unique) >= 8:
break
context_lines, citations = [], []
for i, h in enumerate(unique, start=1):
meta = h.metadata
head = (f"[{i}] {meta.get('title','?')} "
f"(p.{meta.get('page_start')}-{meta.get('page_end')})")
context_lines.append(f"{head}\n{h.text}")
citations.append({
"n": i, "chunk_id": h.chunk_id,
"title": meta.get("title"),
"source_path": meta.get("source_path"),
"page_start": meta.get("page_start"),
"page_end": meta.get("page_end"),
"score": float(h.score),
})
context_block = "\n\n".join(context_lines)
phase_header(5, "Context assembly + answer generation",
f"Top {len(unique)} unique passages β {LLM_CONFIG['model']} via {LLM_CONFIG['provider']}.")
with st.expander("π¦ Context handed to the LLM", expanded=False):
for c in citations:
st.markdown(
f"**[{c['n']}]** {c['title']} Β· pages {c['page_start']}β{c['page_end']} Β· "
f"score `{c['score']:.3f}`"
)
st.code(context_block[:3000] + ("β¦" if len(context_block) > 3000 else ""),
language="text")
t0 = time.time()
ANSWER_SYSTEM = (
"You are a careful research assistant. Use ONLY the provided passages to "
"answer the question. Cite sources inline with [N] where N is the passage "
"number. If the passages are insufficient, say so explicitly."
)
ANSWER_PROMPT = (
f"Question: {query}\n\nPassages:\n{context_block}\n\n"
"Write a concise, well-grounded answer. Use inline citations like [1], [2] "
"that match the passage numbers above."
)
answer = llm.generate(prompt=ANSWER_PROMPT, system=ANSWER_SYSTEM, temperature=0.1)
st.caption(f"LLM generation: {time.time()-t0:.1f} s")
st.markdown("### Answer")
st.markdown(answer)
st.markdown("### Citations")
for c in citations:
st.markdown(
f"**[{c['n']}]** {c['title']} β pages {c['page_start']}β{c['page_end']} "
f"Β· score `{c['score']:.3f}` Β· `{Path(c['source_path']).name}`"
)
# ββ Step 6: critic βββββββββββββββββββββββββββββββββββββ
phase_header(6, "Self-critique",
"LLM scores its own answer for grounding + completeness.")
t0 = time.time()
crit = critique(query, answer, context_block, llm=llm)
c1, c2, c3 = st.columns(3)
c1.metric("Grounded", "β
yes" if crit["grounded"] else "β οΈ no")
c2.metric("Complete", "β
yes" if crit["complete"] else "β οΈ no")
c3.metric("Confidence", f"{crit['confidence']:.2f}",
delta=f"threshold {AGENT_CONFIG['confidence_threshold']:.2f}")
if crit.get("missing"):
st.warning(f"Missing: {crit['missing']}")
st.caption(f"Critique latency: {time.time()-t0:.1f} s")
if crit["confidence"] >= AGENT_CONFIG["confidence_threshold"] and crit["grounded"]:
st.success(f"β Confidence {crit['confidence']:.2f} β₯ threshold β answer accepted.")
if enable_healing:
phase_header(
7, "Self-Healing layer",
"Hallucination detection β chunk quality scoring β knowledge gap β regenerate if needed.",
)
with st.spinner("Running self-healing checksβ¦"):
from healing.healing_loop import self_heal
healed = self_heal(query, answer, unique, citations, llm=llm)
if healed.attempts_used > 0:
st.markdown("### Healed Answer")
st.markdown(healed.answer)
st.markdown("### Updated Citations")
for c in healed.citations:
st.markdown(
f"**[{c['n']}]** {c['title']} β "
f"pages {c.get('page_start')}β{c.get('page_end')} "
f"Β· score `{c['score']:.3f}`"
)
_render_healing_trace(healed.healing_trace, healed.health_score)
return
if it < AGENT_CONFIG["max_iterations"] - 1:
st.warning("Confidence below threshold β refining query and retrying.")
current_query = refine_query(query, crit.get("missing", ""), llm=llm)
else:
st.error("Max iterations reached. Returning best-effort answer.")
if enable_healing:
phase_header(
7, "Self-Healing layer",
"Hallucination detection β chunk quality scoring β knowledge gap β regenerate if needed.",
)
with st.spinner("Running self-healing checksβ¦"):
from healing.healing_loop import self_heal
healed = self_heal(query, answer, unique, citations, llm=llm)
if healed.attempts_used > 0:
st.markdown("### Healed Answer")
st.markdown(healed.answer)
_render_healing_trace(healed.healing_trace, healed.health_score)
# βββββββββββββββββββββββββββββ sidebar + tabs ββββββββββββββββββββββββββββββ
def _sidebar() -> None:
st.sidebar.title("AdaptiveRAG")
st.sidebar.caption("Agentic + Self-RAG + Modular RAG")
llm = _llm()
ok = llm.health()
backend = "Groq API" if HOSTED else "Ollama (local)"
st.sidebar.markdown(f"**LLM backend**: {'π’' if ok else 'π΄'} {backend}")
st.sidebar.markdown(f"**Model**: `{LLM_CONFIG['model']}`")
st.sidebar.markdown(f"**Embedder**: `{EMBEDDING_CONFIG['model'].split('/')[-1]}`")
st.sidebar.markdown(f"**Reranker**: `bge-reranker-base`")
manifest = _load_manifest()
if manifest:
st.sidebar.markdown(f"**Index**: {manifest.get('n_chunks','?')} chunks across "
f"{len(manifest.get('chunks_per_doc',{}))} docs")
with st.sidebar.expander("Documents"):
for doc, n in sorted(manifest.get("chunks_per_doc", {}).items()):
st.markdown(f"- `{doc}` β {n}")
else:
st.sidebar.warning("No index found. Run `python ingest.py --reset`.")
st.sidebar.divider()
st.sidebar.markdown("### Pipeline")
st.sidebar.code(
"question\n β embed (MiniLM)\n β Self-RAG router\n β planner β sub-queries\n"
" β dense β₯ sparse\n β RRF fusion\n β cross-encoder rerank\n β LLM answer\n"
" β self-critique β retry?\n β self-healing βοΈ\n β answer + citations",
language="text",
)
def pipeline_tab() -> None:
st.subheader("π¬ Underhood: watch every stage of the agentic RAG pipeline")
st.caption(
"Each step renders its inputs and outputs as it runs β embedding vector, "
"router decision, planner sub-queries, dense vs sparse hits side-by-side, "
"RRF fusion, cross-encoder rerank, vector-space projection, answer, self-critique."
)
samples = [
"How does Self-RAG decide when to retrieve, and what reflection tokens does it use?",
"Compare DDPM and DDIM sampling β what does DDIM gain by being non-Markovian?",
"What is multi-head self-attention and why does parallelism matter?",
"How does HyDE improve dense retrieval without relevance labels?",
"How does ReAct combine reasoning and acting, vs chain-of-thought?",
"hello, what can you do?",
]
if "vq" not in st.session_state:
st.session_state.vq = samples[0]
cols = st.columns(3)
for i, s in enumerate(samples):
if cols[i % 3].button(s, key=f"vs{i}", use_container_width=True):
st.session_state.vq = s
q = st.text_area("Question", value=st.session_state.vq, height=80, key="vq_input")
enable_healing = st.toggle(
"βοΈ Self-Healing",
value=True,
help="After the answer is generated, run hallucination detection, chunk quality "
"scoring, and knowledge-gap checks β regenerating if issues are found.",
)
if st.button("βΆ Run pipeline", type="primary"):
if q.strip():
visual_pipeline(q.strip(), enable_healing=enable_healing)
def image_tab() -> None:
st.subheader("πΌοΈ Multimodal RAG (Qwen3-VL)")
st.caption(
"Upload an image (e.g. a figure from a paper). Qwen3-VL captions it, the "
"caption + question drives hybrid retrieval, then the model reasons over "
"image + retrieved passages together."
)
uploaded = st.file_uploader("Image", type=["png", "jpg", "jpeg", "webp"])
q = st.text_input("Question about the image", "Explain what this figure shows.")
go = st.button("Reason", type="primary", key="img_go")
if uploaded:
st.image(uploaded, width=400)
if not (go and uploaded):
return
with tempfile.NamedTemporaryFile(suffix=Path(uploaded.name).suffix, delete=False) as f:
f.write(uploaded.getbuffer())
tmp_path = f.name
try:
with st.spinner("Captioning β retrieving β multimodal reasoning..."):
out = image_retrieve_and_reason(tmp_path, q, llm=_llm())
st.markdown("### Caption")
st.write(out["caption"])
st.markdown("### Answer")
st.markdown(out["answer"])
st.markdown("### Retrieved passages")
for i, h in enumerate(out["hits"], start=1):
st.markdown(
f"**[{i}]** {h.metadata.get('title')} "
f"(p.{h.metadata.get('page_start')}β{h.metadata.get('page_end')}) "
f"Β· score `{h.score:.3f}`"
)
st.caption(h.text[:300] + ("β¦" if len(h.text) > 300 else ""))
finally:
os.unlink(tmp_path)
def kb_tab() -> None:
"""Knowledge Base Versioning tab."""
st.subheader("π Knowledge Base Versioning")
st.caption(
"Every ingest run creates a versioned snapshot of the index (kb_v1, kb_v2 β¦). "
"Old snapshots are never deleted β rollback is a single metadata write. "
"You can query any version independently and replay historical answers."
)
# ββ lazy import so the tab loads even if versioning DB is empty ββ
try:
from versioning.version_router import VersionRouter
router = VersionRouter()
history = router.list_versions()
except Exception as exc:
st.error(f"Could not initialise versioning layer: {exc}")
return
# ββ current version banner βββββββββββββββββββββββββββββββββββββββ
current = router.current_version()
if current is None:
st.warning(
"No versioned snapshot found. Run `python ingest.py` to create v1."
)
return
info = router.version_info(current)
c1, c2, c3, c4 = st.columns(4)
c1.metric("Current version", f"v{current}")
c2.metric("Total snapshots", len(history))
if info:
c3.metric("Docs added (latest)", info.get("docs_added", 0))
c4.metric("Docs changed (latest)", info.get("docs_changed", 0))
st.divider()
# ββ version history table ββββββββββββββββββββββββββββββββββββββββ
st.markdown("### Version history")
if history:
df = pd.DataFrame(history)
df["version"] = df["version"].apply(lambda v: f"v{v}")
df["timestamp"] = df["timestamp"].str[:19].str.replace("T", " ")
df = df.rename(columns={
"version": "Version",
"timestamp": "Created",
"batch_name": "Batch",
"docs_added": "Added",
"docs_changed": "Changed",
"docs_unchanged": "Unchanged",
"reason": "Reason",
"collection_name": "Collection",
})
st.dataframe(df, use_container_width=True, hide_index=True)
else:
st.info("No versions yet.")
st.divider()
# ββ rollback βββββββββββββββββββββββββββββββββββββββββββββββββββββ
st.markdown("### Rollback")
st.caption(
"Rolling back points the 'latest' pointer at a previous snapshot. "
"The current snapshot is **not** deleted β you can roll forward again any time."
)
version_nums = sorted([v["version"] for v in history])
if len(version_nums) > 1:
target = st.selectbox(
"Roll back to",
options=[v for v in version_nums if v != current],
format_func=lambda v: f"v{v}",
key="kb_rollback_target",
)
if st.button("βͺ Rollback", type="secondary"):
try:
router.rollback(target)
st.success(f"Rolled back to v{target}. Reload the page to see updated metrics.")
st.rerun()
except Exception as exc:
st.error(str(exc))
else:
st.info("Need at least 2 versions to roll back.")
st.divider()
# ββ per-version diff view βββββββββββββββββββββββββββββββββββββββββ
st.markdown("### What changed per version")
for v in history:
vnum = v["version"]
added = v.get("docs_added", 0)
changed = v.get("docs_changed", 0)
unchanged = v.get("docs_unchanged", 0)
ts = (v.get("timestamp") or "")[:19].replace("T", " ")
badge = "π’" if vnum == current else "βͺ"
label = f"{badge} v{vnum} β {ts} Β· +{added} added, ~{changed} changed, {unchanged} unchanged"
with st.expander(label, expanded=(vnum == current)):
cols = st.columns(4)
cols[0].metric("Added", added)
cols[1].metric("Changed", changed)
cols[2].metric("Unchanged", unchanged)
cols[3].metric("Reason", v.get("reason") or "β")
st.caption(f"Collection: `{v.get('collection_name')}` Β· Batch: `{v.get('batch_name')}`")
# list docs active at this version
try:
from versioning.document_store import DocumentStore
store = DocumentStore()
docs = store.docs_at_version(vnum)
if docs:
st.markdown("**Documents active at this version:**")
for d in sorted(docs, key=lambda x: x["doc_id"]):
status_icon = "β
" if d["status"] == "active" else "ποΈ"
chk = (d.get("checksum") or "")[:12]
st.markdown(
f"{status_icon} `{d['doc_id']}` β "
f"{d.get('title','?')[:60]} "
f"<span style='color:#9aa3b2;font-size:.75rem;'>"
f"sha256:{chk}β¦</span>",
unsafe_allow_html=True,
)
except Exception:
pass
st.divider()
# ββ cross-version query βββββββββββββββββββββββββββββββββββββββββββ
st.markdown("### Query a specific version")
st.caption(
"Run the same question against different snapshots to see how the "
"knowledge base evolution affects retrieval."
)
qv_text = st.text_input(
"Question",
value="How does Self-RAG decide when to retrieve?",
key="kb_query_text",
)
col_v, col_k = st.columns([2, 1])
qv_version = col_v.selectbox(
"Version to query",
options=["latest"] + [f"v{v}" for v in sorted(version_nums, reverse=True)],
key="kb_query_version",
)
qv_k = col_k.slider("Top-K", 3, 10, 5, key="kb_query_k")
if st.button("π Query version", key="kb_query_btn"):
version_arg: str | int = (
"latest"
if qv_version == "latest"
else int(qv_version.lstrip("v"))
)
if not router.collection_exists(version_arg):
st.error(
f"ChromaDB collection for {qv_version} not found. "
"The snapshot may exist in metadata but its collection was removed."
)
else:
with st.spinner(f"Querying {qv_version}β¦"):
try:
hits, resolved = router.query(
qv_text, version=version_arg, k=qv_k, log=True
)
except Exception as exc:
st.error(str(exc))
hits, resolved = [], None
if hits:
st.success(f"Retrieved {len(hits)} passages from v{resolved}.")
for i, h in enumerate(hits, start=1):
meta = h.metadata
title = meta.get("title", "?")
st.markdown(
f"<div class='chunk-card'>"
f"<div class='chunk-meta'>"
f"<span class='pill pill-purple'>#{i}</span>"
f"score <b>{h.score:.3f}</b> Β· {title} Β· "
f"p.{meta.get('page_start')}β{meta.get('page_end')}"
f"</div>{h.text[:300]}{'β¦' if len(h.text)>300 else ''}"
f"</div>",
unsafe_allow_html=True,
)
else:
st.warning("No hits returned.")
st.divider()
# ββ query audit log βββββββββββββββββββββββββββββββββββββββββββββββ
st.markdown("### Query audit log")
st.caption("Every versioned query is recorded here for replay and debugging.")
log = router.get_query_log(limit=20)
if log:
df_log = pd.DataFrame(log)
df_log["timestamp"] = df_log["timestamp"].str[:19].str.replace("T", " ")
df_log["version_used"] = df_log["version_used"].apply(
lambda v: f"v{v}" if v is not None else "β"
)
df_log = df_log.rename(columns={
"timestamp": "Time",
"query": "Query",
"version_used": "Version",
"answer_hash": "Hash",
})
st.dataframe(df_log, use_container_width=True, hide_index=True)
else:
st.info("No queries logged yet.")
def main() -> None:
_sidebar()
st.title("AdaptiveRAG ππ¬")
st.caption(
"Agentic + Self-RAG + Modular RAG over your local paper library β "
f"powered by `{LLM_CONFIG['model']}` via **{LLM_CONFIG['provider']}**. "
"Every pipeline stage is exposed below."
)
pipe, img, kb = st.tabs([
"π¬ Underhood pipeline",
"πΌοΈ Image Q&A (multimodal)",
"π Knowledge Base",
])
with pipe:
pipeline_tab()
with img:
image_tab()
with kb:
kb_tab()
if __name__ == "__main__":
main()
|